Home

Artificial Intelligence - Kenyatta University Library

1. Frame Simba AKO Elephant Colour White 6 All elephants are mammals Frame Elephant AKO Mammal 7 Elephants have a trunk and large ears Frame Elephant AKO Mammal Has parts Trunk Ears Ears Large 8 Ruby is a black and white cat Frame Ruby AKO Cat Colour Black White 9 Ruby is the mother of Pearl and Pearl is the daughter of Ruby Frame Ruby AKO Cat Colour Black White Daughter Pearl Frame Pearl AKO Cat Mother Ruby 10 Some cats are small Frame Unknown cat AKO Cat Size Small 11 All cats eat meat Frame Cat AKO Mammal Food Meat 200 Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises 12 Cats have whiskers and a tail Frame Cat AKO Mammal Food Meat Has parts Whiskers Tail As occurred in Exercise 9 9 1 with production rules and first order logic the frames method of knowledge representation forces the knowledge contained in many of these English sentences to be recast in order to fit in with the method of representation being used This recasting often ends up resulting in subtle changes of the knowledge being expressed or in the meaning not being satisfactorily represented For example
2. Download free ebooks at bookboon com 30 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THE INTERFACE The buttons in the Interface are defined as follows Setup Clears all variables and creates the population of birds at random locations in the environment Step Runs a single tick of the simulation Go Runs the simulation continuously Draw This allows the user to draw collision patches to provide some obstacles for the birds to avoid hitting Erase This allows the user to erase the drawn obstacles Load Image This loads an image from the file named by the Filename slider into the background Although this makes the simulation look more impressive the bird agents cannot see this image If you want to have the birds react to what appears in the image you also need to load the patches for the same image using the Load Patches button Note Since the Load Image button will overwrite patches make sure you synchronise both the Filename and patch name see below so that they match otherwise the birds will react to patches that can t be seen by us as the observers but are there in fact in the environment Load Patches This will load the patches for a given image specified by the patch name slider so that the bird agents are able to see and react to them The sliders in the Interface are defined as follows population This is the number of bird agents created at
3. In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To apply and for all current job openings please visit our web page www ericsson com careers Download free ebooks at bookboon com 44 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour These modified sub behaviours are executed concurrently in no particular order The purpose is to show a Sense amp Think amp Act type behaviour where sensing thinking and acting are done concurrently rather then a Sense Think Act type behaviour where sensing thinking and acting are done one after the other THE INTERFACE The Interface buttons are defined as follows setup This initialises the environment with randomly shaped blocks and places the turtle agents at random locations in the intervening black spaces go This runs the model pen down This gets the turtle agents to put their pen downs thereby drawing the paths they take as they move around The Interface choosers and sliders are defined as follows behaviour Setting this to Standard sets the behaviour to that devised in Uri Wilensky s original model Setting this to Modified sets the behaviour to that explained above in the How it works section choose sub behaviours When the behaviour variable is set to Modified setting th
4. go dijkstra This plots the shortest path distances and find the average degree of separation from the goal node using Dijkstra s algorithm The goal node is specified using the centre of the universe chooser reset blackboards This resets the blackboards that are associated with each node in the network This is used by the Blackboard knowledge sharing mechanism setup walkers This creates new walkers as specified by the number in the slider new walkers per tick These walkers will then start walking around the network trying to find the goal node go walkers once This will make the simulation proceed one step at a time go walkers forever This will make the simulation proceed continuously until it has reached the goal node or it gets stuck reset walkers This kills off all existing searchers and creates new ones to restart the search as specified by the number in the slider new walkers per tick dump blackboards This will dump the contents of all the blackboards in the network this can be used for debugging purposes Download free ebooks at bookboon com 55 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication load blackboards This pre loads the blackboards for each node using the shortest paths found using the Dijkstra algorithm This will improve the performance of the Blackboard knowledge sharing mechanism The model s Interface choosers slid
5. The sequence 2 4 6 8 10 12 14 16 Is The the SOG ENCE of even Whole Numbecs l iS the NUNbEr 00th place in this Sequence Challenging Not challenging Try more gt www alloptions nl life Download free ebooks at bookboon com 32 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THINGS TO TRY What happens if the in cone command is used instead of the in radius command Does substantially reducing the vision cone angle affect the behaviour of the birds EXTENDING THE MODEL Try adding your own images to the model How is the behaviour of the turtle agents affected by the different environments Many of the birds get stuck inside or near to an obstacle This detracts a little from the believability of the simulation Is there some way this bug can be removed from the model NETLOGO FEATURES The model uses the command in radius to implement the boids cone of vision Exercise 6 7 3 Follow and Avoid NetLogo Model Follow and http files bookboon com ai Follow And A void html Avoid W4 gt S ticks 15 30 Setup Go J Adjusts the vision of the boids See eee radius detection 73 aay Boid speed adjusts the step boid speed 5 0 value of each boid per iteration aes speed factor reduces the speed speed factor 3 of both followers and avoiders erry Adds a bit of randomness boid random headi 10 to the boids movements numb
6. sprint length This modifies the length of the paths between the states of the event map network repulsion constant This controls how much each of the states repel each other THINGS TO NOTICE Notice how the repulsion constant slider can be used to repel the states away from each other for larger values and attract the states towards each other for smaller values Notice that the clutter in the network layout can often be removed by setting the value of the spring length slider to zero and then increasing it afterwards THINGS TO TRY Try altering the values of the sliders to see what effect this has on the layout Fast track your career Masters in Management London Business School Regent s Park London NW1 4SA United Kingdom Tel 44 0 20 7000 7573 Email mim london edu www london edu mim Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experience London Business School s Masters in Management will expand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd Applications are now open for entry in September 2011 Fo
7. Download free ebooks at bookboon com 97 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search to implement the classic search algorithms this is a standard solution see for example Russell amp Norvig s AI textbook the model instead adopts a purely agent oriented solution where the information is distributed amongst NetLogo agents These agents conduct the search by transferring information to other agents that continue the search Hence an explicit queue data structure separate from what is happening in the search is not needed although technically the information is still being stored in a queue behind the scenes but the implementation of this is hidden behind the top level commands that NetLogo provides The hope is that this queueless agent oriented approach provides a more intuitive solution that makes it easier to grasp how the search strategies work It uses a situated embodied first person design perspective therefore arguably making it easier to understand the essential differences between the search strategies The breed of searcher turtle agents is used for the searchers that move throughout the network trying to get to the goal These searcher agents maintain information about the current state of the search such as the time taken and the path and estimated costs Each searcher agent expands the search by following all outgoing paths that lead out of the current node
8. ZB ntl 2 twork i d m iM tat Gets iia Nica nodes to add nodes to delete 100 0 emf new walkers per tick 7 time to live Visited Success 53 N A o Succ DOG User Message The average path length to Kevin Bacon is 3 7429 o JE b250 e33C e5706 e220 bO e7100 b34 O bs btg HEE e860 e72 new walkers per node 1 percent in both 7 Figure 7 5 1 Screenshot of the Interface for the Being Kevin Bacon model after the setup network button has been pressed followed by the go dijkstra button to execute Dijsktra s algorithm WHAT IS IT This model applies various algorithms to the problem of searching for a specific goal node in a network The problem is that you do not know where the node is or how the network is configured To solve this problem search needs to be employed to try out the different paths that might lead to the goal node This model implements some multi agent based solutions where knowledge is shared between agents in a dynamic way These can be applied to this problem separately to see how they perform against each other The model uses Kevin Bacon in its title as Kevin Bacon has been chosen as the centre of the Hollywood Universe based on the catch phrase degrees of separation from Kevin Bacon This catch phrase is based on the related mathematical pastime of measuring the degrees of separation of mathematicians from Paul Erd6 s a well known
9. model in the phrase language model which is a mechanism for assigning a probability to a sequence of symbols by means of a probability distribution In the information listed below the two types of models will be explicitly distinguished by using the phrases NetLogo model and language model WHAT IS ITS PURPOSE The purpose of this NetLogo model is to show various important features of language models and to visualise them using NetLogo link and turtle agents HOW IT WORKS The data being modelled is conceptualised as a stream of events i e analogously to Event Stream Processing or ESP where events occur with a specific order for a specific stream but at the same time can occur simultaneously on separate streams A training text sequence is used to train the language model This language model is then used to process a test text sequence Each stream event from the training data is represented by a state turtle agent and paths between states are represented by path link agents When the language model is used for an application such as processing a test sequence of text then walker turtle agents are used to walk the network of states and links to show at what point the current contexts are active Text turtle agents are used for maintaining information about a particular text such as sums and a variable for storing the language model HOW TO USE IT Press the setup button in the Interface to reset all agents To l
10. 7 398 m 7 7 398 4 4 398 v 3 3 398 c 3 3 398 3 3 398 y 3 3 398 p 3 3 398 k 2 2 398 x 1 1 398 2 7398 H O G7 398 O 7398 leti e T Encodings and entropy for next character in sequence 1 2 Cross Entropy for this character 1 bits Total Cross Entropy for all characters so far 44 20078686696105 bits Verve User Message Ax A lt 7 Press OK to continue for next character a Aen i noe i D i Hat G Figure 7 10 1 Screenshot of the Interface for the Shannon Guessing Game model after the load training text and load testing text buttons have been pressed followed by the predict text button and the first 11 characters have been predicted with the slider values as shown in the image ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on making the world a better place for people business and society Join us In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To apply and for all current job openings please visit our web page www ericsson com careers Download free ebooks at bookboon com 77 Artificial Intelligence Exercises Agent Behaviour 7 Communication WHAT IS
11. Do you think the design objectives listed in Section 10 6 of the companion book Artificial Intelligence Agent Behaviour I are good objectives Explain your reasoning Download free ebooks at bookboon com 148 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence 10 7 Towards believable agents Exercise 10 7 1 Chatbot Model Try out the Chatbot model in NetLogo Chatbot http files bookboon com ai Chatbot html chat Then start chatting To begin pr first Ths rules bot setup o begin press setup First This loads the rules Then select who you want to talkto You say Hello Harry says How do you know me Chatbot Harry matched rule no O hello You say Excuse me Harry says I think someone is watching me This line may not be secure Harry failed to match any rule You say Why do you think someone is watching you Harry says Why Is there something wrong with that Chatbot Harry matched rule no 41 why do yout UAS tidsio a Harry 7 debug conversation Figure 10 7 1 Screenshot of the Interface for the Chatbot model after the setup button has been pressed followed by the chat button The chatbot selected by the bot chooser is Harry WHAT IS IT This model implements two basic chatbots Liza and Harry These are named after their more illustrious counterparts Eliza and Parry Unlike Eliza Liza does not try to b
12. If the layout is of type spring then this may help to remove some of the clutter reset searchers This kills off all existing searchers and creates a new one as specified by the start node number Input box to restart the search setup searcher This starts up a new searcher as specified by the start node number Input box this can be used to create multiple searchers at the start of the search go searchers once This will make the search proceed one step at a time go searchers forever This will make the search proceed continuously until it has reached the goal node or it gets stuck The model s Interface choosers sliders switch and Input box are defined as follows network type This selects the type of network that is created when the setup network button is pressed The types of networks are as follows P2P no super nodes This simulates a peer to peer network with no super nodes Download free ebooks at bookboon com 99 Artificial Intelligence Exercises Agent Behaviour 8 Search P2P has super nodes This simulates a peer to peer network with super nodes P2P random single link This simulates a peer to peer network where each node has only one link with another random node P2P incremental This creates the simulated peer to peer network by incrementally building the links to other random nodes one at a time P2P incremental 1 This creates the simulated peer to peer network by increment
13. Regent s Park London NW1 4SA Applications are now open for entry in September 2011 United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Download free ebooks at bookboon com 168 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises NZ xcor warp factor NZ ycor warp factor A straight translation of New Zealand towards Australia on a direct heading does not work the two countries end up hitting each other where their shapes do not fit well together These sliders can be used to warp the shape of New Zealand along the x and y axes respectively so that the shapes more closely fit together when the two countries eventually reach each other Years ago This monitor reports the model s estimate of the number of years ago when the two countries were in the relative positions depicted by the simulation This number is equal to the following ticks x years per tick THINGS TO NOTICE Notice that it is not clear where New Zealand should end up joining with Australia The presence of Tasmania adds an interesting factor in the equation If you run the simulation several times with different slider values you will find out that no simple transformation is possible that will lead to a satisfactory joining up of the two countries Either New Zealand has to be s
14. economics IT or mathematics m www simcorp com S l m Co r p MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 126 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge WHAT IS ITS PURPOSE The purpose of this model is to provide a facility that allows users to draw their own maps thereby allowing them to represent knowledge about a real environment in the virtual 2D NetLogo environment Maps can be considered a form of knowledge representation which provides a different approach with different capabilities for reasoning in comparison to the more traditional forms of knowledge representation such as first order logic and decision trees usually described in AI textbooks A secondary purpose is to show the different ways that objects and paths can be drawn and edited using NetLogo s features HOW IT WORKS The model allows the user to place various objects such as circles squares triangles trees and persons in the environment These objects are represented by different breeds of turtle agents The paths a user can draw on the map have different types The draw path command draws a path using the turtle draw commands i e using the drawing turtle s pen via the pen down command As a consequence this type of path cannot be seen by other agents in the environment so the select region command does not work on this type of path and it cannot be erased using the erase path i
15. or is knowledge processed in a similar manner 9 6 Knowledge without representation Exercise 9 6 1 What do we mean by representation when we talk about knowledge representation Is it fair to say that the ant and termite agents in the Ants and Termites models do not represent knowledge in some manner After all the knowledge is represented globally in the entire multi agent system but not individually within each agent See also Exercise 10 2 1 9 7 Representing knowledge using maps Exercise 9 7 1 Map Drawing NetLogo Model Try out the Map Drawing model in NetLogo Download free ebooks at bookboon com 124 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Legend nean clear background 9 import background end path draw patches pathey draw object erase object select object draw turtles path select region erase region Carousel Figure 9 7 1 1 Screenshot of the Interface for the Map Drawing model with a sample map drawn in the environment WHAT IS IT This model allows the user to draw a map e g A map of Central Park in New York This model was created using NetLogo version 4 0 4 Although it works perfectly fine when converted to NetLogo version 4 1 a run time error occurs when you try to import the Map Drawing Central Park World into the latter version using the CSV file that was exported using the earlier
16. the problems to do with how to represent the meaning of adjectives using production rules and first order logic that arose in Exercise 9 9 1 seem to be more naturally solved using frames although the addition of slot names has added further knowledge that was not originally there in the English sentences such as the Age Gender Colour and Size slot names for examples 1 2 4 5 and 10 Representing the concept all is naturally achieved using inheritance and default values as in examples 6 and 11 but the concept some ends up being problematic as in examples 3 and 10 Also problematic are the examples where there is a verb other than is or are as in examples 3 7 11 and 12 One often used frame based solution for has and have verbs is to use a has parts slot as in the solutions provided It is not clear however what the best way is for representing verbs such as eat as in example 11 Explicit rules as in example 4 are not naturally represented using frames and it is not clear how best to represent such knowledge Possible solutions are to use daemons methods associated with a particular slot or use a different form of inferencing via frame matching or frame inheritance Fast track your career Masters in Management Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experience London Business School s Masters in Management will e
17. 2 1 Whether you agree or disagree with this statement that Knowledge cannot exist by itself independent of us It only exists in our heads depends on your definition of what knowledge is A classical definition of knowledge pre supposes that it is an object in and of itself and therefore can be stored and retrieved in a repository This relates to the classical traditional definition of human memory which relies on common metaphors found in natural language of memory being a place which we store and retrieve information as in a database However an embodied cognitivist perspective would have us believe that knowledge cannot exist outside our own heads Here the important condition is one of embodiment An agent exists as an object embodied within an environment and must interact and react with that environment In G rdenfors conceptual space theory see Section 9 4 each embodied agent has his or her own definitions of concepts he or she believes are justified and relevant and although cultural interactions will ensure that there will be close matches with the concepts held by other agents in that environment there will be subtle differences with no exact match Under this theory then the concepts upon which knowledge arises must necessarily be within our own heads Another way of looking at this is to look at the word knowledge itself The base form of the word contains the word know This is a verb and in common
18. Active Searcher Agents b104 bas 1 1 b3 b63 B85 7 7 b64 b38 Maximum Active Searcher Agents b62 i h67 ase bse b88 b116 b97 354 b72 b73 b8T Total Searcher Agents IDS Depth 0 23 Figure 8 2 3 Screenshot of the Interface for the Searching for Kevin Bacon model for after the setup network button has been pressed followed by the setup searcher button and then the go searchers once button repeatedly with the slider values as shown in the image WHAT IS IT This model applies standard search algorithms to the problem of searching for a specific goal node in a network The problem is that you do not know where the node is or how the network is configured To solve this problem search needs to be employed to try out the different paths that might lead to the goal node This model implements some of the classic search algorithms and these can be applied to this toy problem to see how the different search strategies perform against each other WHAT IS ITS PURPOSE The purpose of this model is to show how some of the classic search algorithms such as breadth first search and depth first search can be implemented in NetLogo and also how they can then be applied to solving a problem in order to compare how they perform HOW IT WORKS The model implements the search algorithms in a novel way by making use of an agent oriented approach which is relatively easy to implement in NetLogo Rather than use a queue data structure
19. BS Y amp C gt X gt V Action Add Z to database Goal Z proven Try this The sequence EE 8 O12 14 1e 1s the SOG ENCE of even Whole Numbecs 00th place in this Orgun Challenging Not challenging Try more WF i options The l 1S the NUNbEr www alloptions nl life Download free ebooks at bookboon com 198 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 9 11 1 There are various ways the slots and values in the frames can be defined for each of these statements Some possible answers are as follows note that AKO stands for A Kind Of 1 Peter is an old man Frame Peter AKO Man Age Old 2 Mary is the sister of Peter Frame Peter AKO Man Age Old Sister Mary Frame Mary Brother Peter 3 Some cats have large ears Frame Unknown cat AKO Cat Ears Large 4 IfX is the sister of Y and Y is a boy then Y is the brother of X Frame Sister AKO Person Gender Female Frame Brother AKO Person Gender Male Frame Girl AKO Person Gender Female Frame Boy AKO Person Gender Male 199 Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises 5 Simba is a white elephant
20. Each of these models have sections in the Information tab that provide various documentation such as what the model is how it works how to use it the meaning of each of the Interface s buttons sliders switches choosers monitors plots and output important things to notice things to try out suggestions for extending the model explanations of interesting NetLogo features used in the model credits and references and links to related models In particular the sections on how to use it things to notice and things to try out provide some suggestions on various things a user can try when playing with the models The reader however should not restrict themselves to just these suggestions Due to the complex system nature of many of the simulations that result from the running of these NetLogo models often unforeseen phenomena emerge as a result of the agent agent and agent environment interactions The reader is encouraged to become an explorer of the virtual environments created by these models by trying out as many of the different combinations of the slider switch and chooser values as possible while running the simulations many times to ensure that a representative sampling of the possible system behaviours is observed Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour Preface This model plays a simple game trying to guess the colour of cars as they drive past It
21. English language the act of knowing is associated with an agent usually a human one We say that we know something not that a car or a tree knows something So knowledge is necessarily associated with us as human agents because it is associated with the process of knowing One further point to note Consider the scenario that the knowledge of how to construct a nuclear bomb has been lost presumably because of a world wide nuclear holocaust Now the classical view would be that this knowledge exists independently of the people who know about it i e what it means and how to apply the knowledge In this scenario then it should be possible then for someone in the future to come along and then find this knowledge again let us say it was written down in a book somewhere and the book was found in some hidden library or archive However the process of recovering the knowledge i e deciphering the meaning of the written text requires that the future agent must develop its own understanding of the problems involved and then the new knowledge is once again in someone s head Notwithstanding this line of reasoning someone could define knowledge as information that can be used to help solve a particular task Under this definition the knowledge as an object argument fits that definition If however you define knowledge as knowing how to use information to help solve a particular task then this definition pre supposes
22. The model s Interface slider and choosers are defined as follows max depth of tree This is the maximum depth of the context tree The order of the model is max depth of tree 1 Download free ebooks at bookboon com 79 Artificial Intelligence Exercises Agent Behaviour 7 Communication output options This determines how much output is written to the output box during the prediction phase which training text This slider allows the user to select from a small but eclectic list of natural language texts to use as the data for training the language model which testing text This slider allows the user to select from a small but eclectic list of natural language texts to use as the data on which the language model is tested by predicting each character one after the other THINGS TO NOTICE Notice how the prediction improves as the max depth of tree variable is increased from 1 to 5 or 6 then drops off slightly However this depends to a great extent on the training text used to build the language model and how closely it is related to the testing text Notice how poor the prediction is when the training and testing texts are from different languages Notice also that for the first few characters in the testing text the prediction is relatively poor Why is this WHAT S MISSING IN THIS EQUATION Please click the advert MAERSK INTERNATIONAL TECHNOLOGY amp SCIENCE PROGRAMME Are you about to g
23. Try this HF Ki options The sequence 2 4 6 8 10 12 14 16 Is The the SOG ENCE of even Whole Numbecs l iS the NUNbEr 00th place in this Sequence Challenging Not challenging Try more gt www alloptions nl life Download free ebooks at bookboon com 98 Artificial Intelligence Exercises Agent Behaviour 8 Search HOW TO USE IT To initialise the search and create a random network whose type is specified by the network type chooser press the setup network button in the Interface To setup a searcher agent to search the network press the reset searchers button first then press the setup searcher button A turtle agent with a shape of a person will be drawn at the node specified in the start node input box To start the search either press the go searchers once button to make the search proceed one step at a time or press the go searchers button to make the search proceed continuously until it reaches the goal node or it gets stuck THE INTERFACE The model s Interface buttons are defined as follows setup network This will clear the environment and all variables and create a random network whose type is specified by the network type chooser Its layout is specified by the layout type chooser The four sliders no of nodes no of super nodes links per node and links per super node control the number of nodes and super nodes and the number of links between them change layout
24. V amp C gt X Rule 4 D gt V Step 3 Sub Goal V from Rule 3 Download free ebooks at bookboon com 196 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Database ABCD Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Step 4 Sub Goal V from Rule 3 Database ABCD Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Action Add V to database Step 2 Sub Goal X from Rule 1 Database ABCDYy Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt xX Rule 4 D gt V Action Add X to database Step 5 Sub Goal Y from Rule 1 Database ABCDYVX Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Step 6 Sub Goal Y from Rule 1 Database ABCDVX Rules base 197 Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises omo A WwW Ne nee 2 V V 2 amp Y amp ADZ amp B gt Y amp C gt X gt V Action Add Y to database Step 1 Sub Goal Z from Rule 1 Database ABCDYV Rules base Rul Rul Rul Rul Oo Oo oO Oo Ls 23 SiH 4 X V V D amp Y amp ADZ
25. a left 90 degrees turn then forward The forward movement behaviour of the searcher agents in the maze is controlled by two choos ers move forward behaviour and the move forward behaviour after turning The options for the type of move forward behaviour used by both these choosers are as follows Move forward n steps unless hit wall The agent moves forward a number of steps determined by the move forward step amount slider Move forward until hit wall The agent will move forward indefinitely until it hits a wall Move forward until any open space The agent will move forward until it has found open space i e until there is a wall in the way or there is open space on both sides Move forward until open space after side wall The agent will move forward until it has found open space i e until there is a wall in the way or there is open space on both sides but will only stop if it has previously been following a wall on either side In Paris or Online International programs taught by professors and professionals from all over the world BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizon
26. a straight line Why is this What settings and type of networks does this occur for Can you get the curve to bend downwards instead of upwards Hint Try changing the value of one or two of the sliders mid stream Notice how the effectiveness of the search deteriorates as the network changes when nodes are either added or deleted Which type of knowledge sharing method seems to cope best with a dynamic network Notice the effect of turning the flag allow revisited states On and Off Why is turning it Of f so effective at dramatically reducing the search for the different search behaviours Notice how the agents shown when the Dijkstra s algorithm is animated seem to jump all around the network Why is this THINGS TO TRY Which knowledge sharing behaviour seems to be the best at this problem Which seems to be the worst Try out each of the different knowledge sharing behaviours to see which one is the most effective Try setting low and high values for the time to live slider How does this affect the percentage of successful searches and the number of nodes visited as a result Try pressing the setup walkers button multiple times This will start the search with more than one walker Try out all the different types of networks with different slider settings Which types of network causes problems for the various knowledge sharing methods and which do not i e Which seem to be more difficult to search Try switching the knowl
27. all boids Clear This clears the trails The model s Interface sliders switches and chooser are defined as follows 1 2 For initialising the simulation population This specifies the number of boids turtle size This specifies the size of the boids labelID If this is set to On then the boids are labelled with their who numbers For controlling the movement of the boids rate of random turn This controls how much the wandering boid turns each time tick The boid has a tendency to head in a right turning direction as the rate of random turn to the right as specified by the slider is twice that of the rate of random turn to the left boid speed This controls the speed of the boid i e how much it moves forward each tick For specifying parameters that control boid collision avoidance behaviour collision This turns on collision detection and avoidance if set to On turtle radius angle This defines the radius angle of the boid s vision cone in relation to other boids turtle radius length This defines the radius length of the boid s vision cone in relation to other boids 4 For specifying parameters that control patch collision avoidance behaviour where the patches are the boundaries of the path radius angle This defines the radius angle of the boid s vision cone in relation to the edges of the path radius length This defines the radius length of the boid s vision cone in relation to the edges of th
28. an agent is involved who must define what it means to know something and what information is meaningful Download free ebooks at bookboon com 185 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 9 4 2 Below is a short answer to this question A full answer should provide the gist of the following answer plus also demonstrate through examples an understanding of the fundamental differences in the three different approaches In all three approaches there are problems in describing the knowledge required to describe each of the five examples Note that the five senses are all represented here hearing taste feeling smell and sight Rather than dealing with the problems with each sense it is simpler to deal with the overall problem of sensing Sensing and perception is necessarily associated with an agent embodied in an environment Hence to describe the knowledge requires a cognitivist approach In that light conceptual spaces theory offers the best tool to give us some leverage into what is essentially a very difficult problem There are major problems with the symbolic layer not least to do with the issues to do with embodiment i e The existence of the agent within an environment and understanding of the interaction of the agent with everything that exists within it and being able to perceive and sense changes of the environment R
29. around one of the obstacles THINGS TO TRY Try adjusting the boid s speed radius angle and radius length to see how this affects the boid s behaviour Also try changing the Interface Settings to see if this has any affect Try adding obstacles to see how this affects the boid s ability to cover the entire environment For example add obstacles in the form of a maze Try to create black spots where the boid never visits Alternatively try to trap the boid into a small area EXTENDING THE MODEL The model could be extended to add gradual acceleration and deceleration This would enhance the simulation of the boids model Download free ebooks at bookboon com 42 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour NETLOGO FEATURES The code uses the in cone command to simulate the boid s cone of vision RELATED MODELS See the following models Crowd Path Following Flocking With Obstacles Follow and Avoid Obstacle Avoidance 1 Obstacle Avoidance 2 Vision Cone Example 2 Wall Following Example 2 These are basic implementations of various Craig Reynold s steering behaviours for boids Another boid related model is the Biology Flocking model in the Models Library Exercise 6 7 6 Wall Following Example 2 NetLogo Model WS ticks 538 setup go g Fs pen down blue follow right hand wall green follow left hand wall behaviour Modified v choose sub behaviours hoose
30. at node n4 Expanding the search for searcher 14 by hatching new searcher 28 at node n4 Expanding the search for searcher 14 by hatching new searcher 29 at node n8 Expanding the search for searcher 17 by hatching new searcher 30 at node n2 Download free ebooks at bookboon com 178 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Expanding the search for searcher 17 by hatching new searcher 31 at node n4 Expanding the search for searcher 17 by hatching new searcher 32 at node n6 Expanding the search for searcher 17 by hatching new searcher 33 at node n8 Expanding the search for searcher 32 by hatching new searcher 34 at node nO Searcher agent 34 has found the goal node nO The output produced for the Depth First Search on Network 1 is as follows Figure 8 3 1 2 Screenshot of the Environment for the Searching for Kevin Bacon 2 model showing how the Depth First Search searches the network shown loaded from the Network 1 data file The searcher agent 9 is starting at node 1 Expanding the search for searcher 9 by hatching new searcher 10 at node n2 Expanding the search for searcher 9 by hatching new searcher 11 at node n4 Expanding the search for searcher 10 by hatching new searcher 12 at node n1 Expanding the search for searcher 10 by hatching new searcher 13 at node n3 Expanding the search for searcher 10 by hatching new searcher 14 at node n5 Expanding the search for searcher 12 by
31. can be used to prune out what is visualised of each language model Only the specified streams will be visualised pick these events This can be used to prune out what is visualised of each language model Only the specified events will be visualised test text This is the testing text that is being processed sequentially Total of paths traversed plot This plots the number of paths traversed in the language model versus the number of events processed Most Active Recently in Window This is the most active model in the past window of length window size in the testing text THINGS TO NOTICE Notice that word based language models are significantly smaller in complexity when visualised than the complexity of the character language models This is to be expected since there are a lot less word events in a text than character events for example there are on average for the English language 5 characters per word Try loading several language models all together Then notice how when a testing text is chosen with language similar to one of the training texts used to train the language models then the language model that becomes most active is usually the one with the similar language This is because language can readily be identified uniquely and explains why it is possible to do authorship ascription using language models with a high degree of accuracy Download free ebooks at bookboon com 75 Artificial Intelligence
32. created and randomly located through the network The setup walkers button can be pressed multiple times if you wish more walkers to start the search To start the simulation either press the go walkers once button to make the search proceed one step at a time or press the go walkers button to make the search proceed continuously The walkers will continue walking the network until they find the goal node or they expire without success if they have exceeded the time to live slider value The links that the walker agents Download free ebooks at bookboon com 54 Artificial Intelligence Exercises Agent Behaviour 7 Communication traverse across are briefly thickened to show the paths the agents are taking as they proceed in their search To plot the shortest path distances and find the average degree of separation from the goal node using Dijkstra s algorithm press the go dijkstra button THE INTERFACE The model s Interface buttons are defined as follows setup network This will clear the environment and all variables and create a random network whose type is specified by the net work type chooser Its layout is specified by the layout type chooser The four sliders no of nodes no of super nodes links per node and links per super node control the number of nodes and super nodes and the number of links between them change layout If the layout is of type spring then this may help to remove some of the clutter
33. criteria time complexity space complexity optimality and complexity Relate the theory to what happens in practice Try out all the different types of networks with different slider settings Which types of network cause problems for the various searches and which do not 1 e Which seem to be more difficult to search Try switching the searching behaviour mid stream For example try switching from a breadth first search to a depth first search then back again Try out the different heuristics for the informed searches Greedy Best First Search and A Search Do they have any effect on how effective the search is EXTENDING THE MODEL Try adding other search algorithms or adding your own variations to the ones implemented in the model Try adding your own type of network either randomly created or with a correspondence to a network in real life You will need to add input routines to create the network layout for the latter NETLOGO FEATURES Note the use of the hatch and die commands to clone child searchers and terminate parent searchers thus avoiding the need for a separate queue data structure to maintain the information Download free ebooks at bookboon com 102 Artificial Intelligence Exercises Agent Behaviour 8 Search during the search Note also the variation in the ask command that determines the strategy being used that defines the different searches RELATED MODELS For other search problems see
34. defined as follows setup This clears the environment and loads a new image as specified by the map or image to use chooser draw arrow This will draw an arrow over the top of the image If the user then clicks and drags the mouse an arrow is drawn with the tail at the point the mouse was clicked and the head at the point the mouse was released The mouse must be both clicked and dragged otherwise no arrow is drawn The colour and dimensions of the arrow are specified by the arrow colour arrow start size arrow end size and arrow thickness sliders To get just a single arrow head shape without the shaft and arrow tail then use the erase object button to erase the end of the arrow This can also be used to erase the arrow head and leave just the tail if so desired draw circle This will draw a circle over the top of the image at the point the mouse is next clicked The colour and dimensions of the circle are specified by the circle colour circle size and circle thickness sliders g and automatic nomical to create rience expertise and creativity ost performance beyond expectations eed the best employees who can challenge The Power of Knowledge Engineering q A Plug into The Power of Knowledge Engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 189 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises draw text This will d
35. gt Carnivore x 12 Cats have whiskers Rules database IF x is cat Vx Cat x Has_whiskers x A and a tail THEN x has whiskers AND x Has tail x has tail Despite the relative simplicity of these examples i e considering the text in the English sentences many of these pose problems for the production rules representation and or the first order logic representation Where there is a straight relationship being expressed between nouns such as in examples 2 4 6 and 9 then the representations seem well suited to expressing the knowledge However when there is an adjective modifying a noun as in examples 1 3 5 7 8 and 10 then the adjective in the rules or logic statements plays the same role as the noun i e Cat Ruby Black Ruby rather than modifying it and in so doing this describes a slightly different relationship than was expressed in the original text Similarly when there is a verb other than the verbs is or are in Download free ebooks at bookboon com 193 Artificial Intelligence Exercises Agent Behaviour the text as in the examples 3 7 11 and 12 then both the production rules and first order logic statements have to use a contrived mechanism for expressing the knowledge for example eats meat Eats_meat and has whiskers Has_whiskers Alternative representations in these cases also end up being less than satisfactory by not completely ca
36. hatching new searcher 15 at node n2 Expanding the search for searcher 12 by hatching new searcher 16 at node n4 Expanding the search for searcher 15 by hatching new searcher 17 at node n1 Expanding the search for searcher 15 by hatching new searcher 18 at node n3 Expanding the search for searcher 15 by hatching new searcher 19 at node n5 Expanding the search for searcher 18 by hatching new searcher 20 at node n2 Expanding the search for searcher 18 by hatching new searcher 21 at node n6 Expanding the search for searcher 21 by hatching new searcher 22 at node nO Searcher agent 22 has found the goal node n0 The output produced for the Greedy Best First Search on Network 1 is as follows Download free ebooks at bookboon com 179 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Figure 8 3 1 3 Screenshot of the Environment for the Searching for Kevin Bacon 2 model showing how the Greedy Best First Search using the heuristic Manhattan Distance searches the network shown loaded from the Network 1 data file The searcher agent 9 is starting at node 1 Expanding the search for searcher 9 by hatching new searcher 10 at node n2 Path cost of searcher 10 20 Expanding the search for searcher 9 by hatching new searcher 11 at node n4 Path cost of searcher 11 10 WHAT S MISSING IN THIS EQUATION Please click the advert You could be one of our future talents MAERSK INTERNATIONAL T
37. html Try this HF 1 options The Sequence 9A amp 8 10 12 14 16 Is the SOG vice of even whole Numbecs The 00th place in this sequence S the number Challenging Not challenging Try more gt www alloptions nl life Download free ebooks at bookboon com 165 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises nos P Yt p Years ago 222600000 Figure S 6 3 2 Screenshot of the Interface for the ANZ Continental Drift model after the setup button has been pressed followed by the continental drift button with the slider values as shown in the image WHAT IS IT This model shifts New Zealand back towards Australia in order to illustrate the process of continental drift In effect the model is running time backwards in order to show where New Zealand was in relation to Australia millions of years ago WHAT IS ITS PURPOSE The main purpose of the model is not to accurately simulate what has occurred during the continental drift process between Australia and New Zealand over hundreds of millions of years since that would require much more code well grounded in scientific evidence than what is used in this model Instead the primary aims of the model are to provide a simple animation of the concept of continental drift to allow the user to try out different possibilities to see where Australia and New Zealan
38. in the entire multi agent system but not individually within each agent See also Exercise 9 6 1 Download free ebooks at bookboon com 147 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence 10 3 What Al can and can t do Exercise 10 3 1 Write a list of 10 tasks that humans can do that require some degree of intelligence Find out whether there are computer systems that can do each of the tasks on your list How well do they perform at each of the tasks Are they capable of successfully completing the task What human level of performance are they capable of novice intermediate or expert 10 4 The Need for Design Objectives for Artificial Intelligence Exercise 10 4 1 Research in the A I literature how the term intelligence is being defined Are there distinctly different meanings of the term or do they all spring from the same generic meaning Do the publications provide any justification for why the systems described can be called intelligent 10 5 What are Good Objectives Exercise 10 5 1 Devise your own design objectives for Artificial Intelligence Become your own worst critic analyse your objectives to see if they are SMARTER Which attributes did you fail on i e are they specific enough are they measurable are they achievable are they realistic are they time bound and so on 10 6 Some Design Objectives for Artificial Intelligence Exercise 10 6 1
39. of knowledge representation as shown by this model and also the associated reasoning processes such as forward and backward reasoning for rules and frame based reasoning using frame matching methods They also provide a useful way of visualising the different types of knowledge representation and of animating the reasoning processes HOW TO USE IT First pick a type of knowledge representation using the KR t ype chooser in the Interface Then pick a problem using the select problem chooser To load the knowledge related to the problem press the setup knowledge button This will display the knowledge in the large black environment box on the left If the visualisation is a bit cluttered at first and the Layout type has been set to spring press the change layout button while playing around with the spring length spring constant and repulsion constant sliders Then make sure the reasoning process is reset to the initial state by pressing the rest reasoning button Finally to perform the reasoning process press the go reason button THE INTERFACE The model s Interface buttons are defined as follows setup knowledge This loads the knowledge for the problem as specified by the select problem chooser using the knowledge representation specified by the KR type chooser The knowledge is visualised in the environment using an event map finite state automata that depicts the relationships between knowledge states dump knowledge This
40. of their node number For example expand node n1 before every other node expand node n2 before every other node except node n1 and so on Also allow nodes to be revisited Use Manhattan Distance as the heuristic for the problem Download free ebooks at bookboon com 113 Artificial Intelligence Exercises Agent Behaviour 8 Search Compare how the informed search algorithms Greedy Best First Search and A Search perform on the same problem WHAT S MISSING IN THIS EQUATION You could be one of our future talents Please click the advert MAERSK INTERNATIONAL TECHNOLOGY amp SCIENCE PROGRAMME Are you about to graduate as an engineer or geoscientist Or have you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour 8 Search 8 8 Comparing the search behaviours Exercise 8 8 1 Have a play with the Searching Mazes Missionaries and Cannibals and Searching for Kevin Bacon NetLogo models to try out the different search algorithms that have been implemented in each model e Breadth First Search e Uniform Cost Search e Depth First Search e Multi agent Depth First Search e Depth Limited Search e Iterative Deepening Search e Greedy Best First Search e A Search e Hill Climbing Search Try different settings to see what ha
41. one at ra Y SS S number of turtles 60 Figure 6 7 6 Screenshot of the Interface for the Wall Following Example 2 model after the setup button has been pressed followed by the go and pen down buttons Download free ebooks at bookboon com 43 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour WHAT IS IT The turtles in this example follow walls made out of colored patches The blue turtles try to keep the wall on their right the green turtles keep the wall on their left Hence the blue turtles end up heading in a clockwise direction and the blue turtles end up in an anti clockwise direction Only the Information in the model that is different to the Wall Following Example model provided by Uri Wilensky is included below HOW IT WORKS The model implements the standard unmodified behaviour as provided by Uri Wilensky s original model but also provides an alternative modified behaviour where the actions have been split into three independent parts as follows 1 walk modified 1 The turtle turns right if necessary 2 walk modified 2 The turtle turns left if necessary 3 walk modified 3 The turtle moves forward ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on making the world a better place for people business and society Join us
42. people those that spread messages and those that don t Try a further variation where there is a random probability as defined by a slider value that the message may mutate when it is passed on Set various conditions such as how many times does the message need to be heard from different people that determine when the message is finally believed How long does it now take before the message is spread 7 5 The Small World Phenomenon and Dijkstra s algorithm Exercise 7 5 1 Being Kevin Bacon NetLogo Model Try out the Being Kevin Bacon model in NetLogo Download free ebooks at bookboon com 51 Artificial Intelligence Exercises Agent Behaviour Us ticks 0 Number of walkers setup network setup walkers 10 change layout go walkers once at EY NR ee baa E x a go dijkstra go walkers g ai 2 b89 0 reset blackboards reset walkers 0 Tick 10 HOn yon z b30 io animate dijkstra io revisits allowed 7 p Successful searches 590 k e centre of the universe knowledge sharing b58 p NI A Kevin Bacon v Blackboard v N Xeb78 network type layout type i b12 283 eos P2P has super n y spring y j TR aT x i 0 Searches 10 E jee _ Shortest path dista no of nodes 100 no of super nodes 5 e98 i ji i s l li 1 links per node 4 links per super node 20 4 l NI j i 5 a Al dump blackboards load blackboards ot memes eee
43. performed best on the various network types Also try to spot any interesting results or trends in the data produced Create further experiments using different settings when the Experiment dialog box appears Did these experiments produce different results to the first experiment 7 7 Entropy and Information Exercise 7 7 1 Work out the entropy for messages to be encoded using symbols from the following probability distributions e foran alphabet containing two symbols with equal probabilities e foran alphabet containing four symbols with equal probabilities e for an alphabet containing six symbols Red Orange Brown Yellow Green and Lime with probabilities as follows p Red p Orange p Brown a p Yellow gt p Green 5 p Lime Exercise 7 7 2 Under what conditions is the entropy minimum for an alphabet of a specific size 7 8 Calculating Entropy in NetLogo Exercise 7 8 1 Cars Guessing Game NetLogo Model Try out the Cars Guessing Game model in NetLogo Download free ebooks at bookboon com 66 Artificial Intelligence Exercises Agent Behaviour 7 Communication setup animate cars a Adaptive Agent s distribution Daban wick adaptive white Source Entropy Source Code Length 4 6E12 H source Source distribution Agent 01 s distribution 5 0 52 162 57 i H Adaptive H Agent 01 adaptive black Adaptive Entropy Adaptive Code Length 2 H Agent 02 9 0 57 180 96 4 H Age
44. procedure is called In Paris or Online International programs taught by professors and professionals from all over the world BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizonsuniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 20 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Exercise 6 6 2 Wall Following Events NetLogo Model Try out the Wall Following Events model in NetLogo Wall Following Events http files bookboon com ai Wall Following Events html dS ticks 0 setup change layoutg er spring con 0 05 Ss il spring length 42 CE repulsion c 100 Figure 6 6 2 Screenshot of the Interface for the Wall Following Events model after the setup button has been pressed followed by the change layout button with the slider values as shown in the image WHAT IS IT This model visualises a small set of events that an agent can follow in order to perform a modified type of wall following behaviour where sensing thinking and acting are all done concurrently in no particular order see the Wall Following Examp
45. randomly in the network heuristic This specifies the heuristic to be used for the Informed searches Greedy Best First Search and A Search max agents to expand This sets the maximum number of agents to expand the search each step It is only used by the Multi Agent Depth First Search Download free ebooks at bookboon com 100 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search allow revisited states This is a flag that if set to On will allow the search to proceed to already visited states max depth This sets the maximum depth of the search for the depth limited search and the Iterative Deepening Search only The model s Interface monitors are defined as follows Active Searcher Agents This shows the current number of active searcher agents still participating in the search Maximum Active Searcher Agents This shows the maximum value that the Active Searcher Agents monitor has achieved throughout the search Total Searcher Agents This shows the total number of searcher agents that the search has used IDS Depth This is the current depth during the execution of the Iterative Deepening Search Fast track your career Masters in Management Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experience London Business School s Masters in Management will expand your thinking and provide you
46. remains burning once it has been ignited and has already spread the fire to its neighboring trees THINGS TO NOTICE Notice how much of the forest burns for different density settings What threshold does the density setting need to be reached so that almost all of the forest gets burned Notice what happens when the unburnt forest is used to reset the forest back to the way it was when the setup button was pressed and then the go button is pressed again Are the same parts of the forest burnt If so then why Notice that the turtles do not move How can you check whether this is true But yet the fire seems to move How Notice which arrangements of firebreaks seem to be best at containing the fire Notice which parts of the imported satellite image are confused with trees How could the is tree reporter be improved to make a better categorisation of the image Download free ebooks at bookboon com 161 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises THINGS TO TRY Try drawing firebreaks as the fire is spreading Often it is very difficult to contain the fire and prevent it from spreading further Slowing down the simulation using the Interface speed slider can help but this will also slow down the drawing of the firebreaks Try drawing a firebreak from one side of the environment to the other Then try leaving in some gaps Also try drawing firebreak circl
47. that particular sense well Solution to Exercise 9 7 2 Map and Image Annotator Model The Map and Image Annotator NetLogo model provides one solution to this problem WS ticks 0 Bethesda Terrace draw circle erase object al select region erase region 2 Colours are as follows red 15 orange 25 yellow 45 green 55 blue 105 magenta 125 pink 135 Here is Mark Figure 9 7 2 Screenshot of the Interface for the Map and Image Annotator model with a sample image imported into the environment and sample annotation Download free ebooks at bookboon com 187 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises WHAT IS IT This model allows the user to annotate a map or an image The model imports an image into the background for the environment The user can then add arrows circles text and marks i e x s These can be later edited or erased if required HOW IT WORKS The map or image is placed in the background using the import drawing command Separate turtle agent breeds arrow circle text and mark are used for drawing the corresponding object in the environment A further side turtle agent is used for drawing the sides of the rectangular box used by the select region and erase region buttons for selecting and erasing multiple objects Arrows are drawn using two arrow turtle agents The start a
48. the ABBA song The Winner Takes It All The gods may throw a dice Their minds as cold as ice And someone way down here Loses someone dear The winner takes it all The loser has to fall It s simple and it s plain Why should I complain The judges will decide The likes of me abide Download free ebooks at bookboon com 81 Artificial Intelligence Exercises Agent Behaviour 7 Communication Spectators of the show Always staying low The game is on again A lover or a friend A big thing or a small The winner takes it all Exercise 7 11 2 How might it be possible to communicate with an alien race Is there some common frame of reference with which we might communicate Could mathematics for example provide a common ground to facilitate communication But is human mathematics itself based too much on our own human embodied existence Is there an alternative approach where we could encode meaning directly in the message we wish to convey to the alien race Exercise 7 11 3 Analyse the following excerpt from Bill Gates autobiography My Life Al Gore gave a marvelous concession speech It was genuine gracious and patriotic When I called to congratulate him he told me that a friend who was a professional comedian had joked to him that he had gotten the best of both worlds he had won the popular vote and didn t have to do the job The next morning after Tony Blair and I talked a bit I walked o
49. the Being Kevin Bacon model the Missionaries and Cannibals model and the Searching Mazes model For an insight into the Manhattan distance and Euclidean distance metrics used as heuristics for the Informed searches see the Manhattan Distance model Exercise 8 2 4 Searching for Kevin Bacon 2 NetLogo Model Try out the Searching for Kevin Bacon 2 model in NetLogo Searching for Kevin Bacon 2 http files bookboon com ai Searching for K evin Bacon 2 html a Heine setup network setup searcher reset searchers lA ticks 3 3 change layout x go searchers once go searchers 2 network type layout type search behaviour P2P no supe Y spring v A Search v heuristic Set start node start node number Euclidean distance v number to 0 for random start 7 node size 1 5 no of nodes 8 no of super nodes 1 ee 5 links per super node 42 ye _ spring len 12 aE 40 on allow revisited nodes B specify coordinates Ta output trace IDS Depth Active Searcher Agents 3 0 Total Searcher Agents Maximum Active Searcher Agents 6 3 Expanding the search for searcher 8 by hatching new searcher 9 at node n6 Path cost estimated cost combined cost of searcher 9 12 78 27 46 40 24 Expanding the search for searcher 9 by hatching new searcher 10 at node n2 Path cost estimated cost combined cost of searcher 10 25 55 35 03 60 59 0 Expanding
50. the environment and therefore to keep the same aspect ratio there will be some unfilled areas in the environment as a result Download free ebooks at bookboon com 191 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises THINGS TO TRY Try using the model to annotate your own map or image such as a photo you might have taken on holiday Then use the print screen feature on your computer to save the annotated image EXTENDING THE MODEL Try modifying the model so that it allows the user to draw lines and add further objects and NETLOGO FEATURES The code makes use of the watch subject and perspective commands when the select object button is used to place a halo around the selected object RELATED MODELS See the Mouse Drag Multiple Example and Mouse Drag One Example models CREDITS AND REFERENCES The model makes use of code provided by Uri Wilensky in the Mouse Drag Multiple Example and Mouse Drag One Example models The financial industry needs a strong software platform That s why we need you Find your next challenge at www simcorp com careers m www simcorp com S l m Co rp MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 192 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 9 9 1 Sentence Production R
51. the search for searcher 9 by hatching new searcher 11 at node n3 Path cost estimated cost combined cost of searcher 11 29 24 13 15 42 39 i load network from a file save network to a file Expanding the search for searcher 9 by hatching new searcher 12 at node n7 M generate problem description add to network Figure 8 2 4 Screenshot of the Interface for the Searching for Kevin Bacon 2 model for after the setup network button has been pressed followed by the setup searcher and go searchers buttons with the slider values as shown in the image The screenshot shows a small network with 8 nodes with the Output box showing a trace of the search The Information for this model is the same as the Information for the Searching for Kevin Bacon model detailed in Exercise 8 2 3 except for the following Download free ebooks at bookboon com 103 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search WHAT IS IT This NetLogo model is an extension to the Searching for Kevin Bacon model that has an Output that provides an English description of the search problem and provides a trace of how the search proceeds Specific networks can also be saved to a file and loaded latter into the application As with the Searching for Kevin Bacon model this model applies standard search algorithms to the problem of searching for a specific goal node in a network HOW TO USE IT To
52. the window size over which the NetLogo model tries to classify the text using a crude form of text categorisation In this case the activity count of the language models is maintained where an activity is the highlighting of a crossed link during the animation and the most active language model is shown in the Most Active Recently in Window monitor The category of the testing sequence can then be set to the language label of the most active language model i e If there were French English and Spanish language models loaded from various training texts and the most active language model shown by the monitor was the French language model then the text can be categorized as being French 6007 A a sways is Wy 282 aus It s only an opportunity if you act on it IKEA SE STUDENT ev oh a i Download free ebooks at bookboon com 74 Artificial Intelligence Exercises Agent Behaviour 7 Communication The NetLogo model s Interface switches are defined as follows setup char models If set to On this sets up the models to recognize words and punctuation from characters show counts If set to On this will show the counts in the visualisation of the language models debug trace If set to On this will print out copious amounts of debug information to show the code being executed The NetLogo model s Interface input boxes plot and monitor are defined as follows pick these streams This
53. three models such as Search is completed No more searchers Abort Made it to the centre of the maze and Found Kevin Bacon replace them with reporters and use these to specify the Stop condition field for the experiment For the Searching Mazes model and the Missionaries and Cannibals model use the setup and go procedures to fill in the Setup commands and Go commands fields For the Searching for Kevin Bacon model use the setup network and go searchers procedures as the setup and go commands instead Download free ebooks at bookboon com 116 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge 9 Knowledge 9 1 Knowledge and Knowledge based Systems Exercise 9 1 1 As pointed out in Exercise 1 1 1 of the book Artificial Intelligence Agents and Environments one of the issues for Artificial Intelligence research concerns the problem of categorization for intelligent systems the problem of how to select suitable categories to cover a set of examples and the resultant classification errors that arise once a particular set of categories has been chosen As a further example use your favourite search engine to find out how the terms Knowledge and Knowledge based System are defined by different people Classify the various definitions you find into a set of different categories Which definitions fit well into your set of categories Which definit
54. type slider is set to spring and can be used to control how the network looks For example setting the spring length to 0 then increasing it will often unclutter the layout of the network Download free ebooks at bookboon com 105 Artificial Intelligence Exercises Agent Behaviour 8 Search The model s Interface switches are defined as follows specify coordinates If this switch is set to On when the network type slider is set to User specified then the user has to specify the precise co ordinates of each node rather than using the mouse to point and click on a location output trace If this switch is set to On then a trace of the search is written to the Output box this is located to the middle right at the bottom of the Interface THINGS TO TRY Try to follow how the informed searches Greedy Best First and A search the various networks Trace the searches as they proceed by turning the output trace switch to On Work out how the path cost and estimated cost are calculated at each stage of the search Try out the different heuristics Do they have any effect on how effective the search is RELATED MODELS For other search problems see the Being Kevin Bacon model the Missionaries and Cannibals model Searching for Kevin Bacon and the Searching Mazes model For an insight into the Manhattan distance and Euclidean distance metrics used as heuristics for the Informed searches see the Manhattan Distance mod
55. using a halo the object can be moved around using the mouse The select region button allows the user to select a region of objects to be moved around using the mouse or erased using the erase region button If you wish to draw a path there are three options paths drawn using a turtle agent s pen paths drawn using patch agents and paths drawn using turtle agents The first option comes out looking smoother the other two options can end up being patchier When drawing the path move the mouse to the desired location and then click and drag to draw the path To draw a path using a turtle agent s pen first choose the desired path colour and path size then press the draw path button The path cannot be erased once it has been ended either when end path has been pressed or another object or path has been drawn elsewhere If this is not the case then erase path will work by erasing each point in the path one by one If erase path is not successful at deleting the just drawn path then the only way to remove it is to draw over the top of it To draw a path using patch agents select the desired patches path size then press the draw patches path button To draw a path using turtle agents select the desired turtles path size then press the draw turtles path button Text can be added by the draw text button Before doing so make sure that you type in the text you want to draw into the text input box first then select the desired text col
56. with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School Regent s Park London NW1 4SA United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Applications are now open for entry in September 2011 Download free ebooks at bookboon com 101 Artificial Intelligence Exercises Agent Behaviour 8 Search THINGS TO NOTICE Notice the effect of turning the flag allow revisited states On and Off Why is turning it Of f so effective at dramatically reducing the search for the different search behaviours Notice that some of the searches if repeated with the same settings end up in different places Why is this Notice that many of the behaviours result in the searching turtle agents getting stuck When does this happen and why THINGS TO TRY Which search behaviour seems to be the best at this problem Which seems to be the worst Try out each of the different search behaviours to see which one is the most effective Find out how well each search algorithm performs against the four evaluation
57. 0 at node n2 Path cost estimated cost combined cost of searcher 10 20 40 60 Expanding the search for searcher 9 by hatching new searcher 11 at node n4 Path cost estimated cost combined cost of searcher 11 10 50 60 Expanding the search for searcher 11 by hatching new searcher 12 at node n1 Path cost estimated cost combined cost of searcher 12 20 60 80 Expanding the search for searcher 11 by hatching new searcher 13 at node n5 Path cost estimated cost combined cost of searcher 13 30 30 60 Expanding the search for searcher 11 by hatching new searcher 14 at node n7 Download free ebooks at bookboon com 181 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Path cost estimated cost combined cost of searcher 14 20 40 60 Expanding the search for searcher 13 by hatching new searcher 15 at node n2 Path cost estimated cost combined cost of searcher 15 40 40 80 Expanding the search for searcher 13 by hatching new searcher 16 at node n4 Path cost estimated cost combined cost of searcher 16 50 50 100 Expanding the search for searcher 13 by hatching new searcher 17 at node n6 Path cost estimated cost combined cost of searcher 17 50 10 60 Expanding the search for searcher 13 by hatching new searcher 18 at node n8 Path cost estimated cost combined cost of searcher 18 40 20 60 Expanding the search for searcher 18 by hatching
58. 50 degrees ess eee radius length 20 length Draw Obstacle wy bad Follow Wanderer Figure 6 7 4 Screenshot of the Interface for the Obstacle Avoidance 1 model after the setup button has been pressed followed by the go button WHAT IS IT This model implements a boid see Craig Reynold s work that employs basic obstacle avoidance steering behaviour Download free ebooks at bookboon com 36 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour HOW IT WORKS It does this by generating a wanderer turtle boid that simply wanders around randomly in the environment avoiding the obstacles that are drawn with white patches The boid is implemented using NetLogo s in cone command that implements a turtle with a cone of vision INTERFACE The model s Interface buttons are defined as follows Setup This sets up the environment with a tree like arrangement of obstacles drawn with white patches One turtle agent the wanderer boid is created and placed at a random location Go The boid starts wandering around the environment avoiding obstacles Draw Obstacle The user can draw further obstacles in the environment These are also coloured white Follow Wanderer This allows the perspective of the visualisation to be altered so that it is centred on the wanderer The model s Interface sliders are defined as follows boid speed This controls the speed of the boid i e how much i
59. 70 Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises The Information for this model is the same as that provided by the State Machine Example model except for the following WHAT IS IT This model has been slightly modified by Bill Teahan from the State Machine Example model in NetLogo s Models Library It has added a plot to graph the growth in the number of piles or equivalently the number of spaces during the simulation This is in order to estimate the progress of the self organisation of the system as a whole WHAT IS ITS PURPOSE The purpose of the modified model is to ascertain whether a function based on global features can be used to measure the state of the simulation i e how self organised the system is If the graph of the function shows a smooth upward curve then it might be useful for example as a fitness function for an evolutionary algorithm in order to evolve the behaviour of the termites automatically rather than being explicitly programmed HOW TO USE IT First select a function to plot using the pileness measure chooser Then press the setup button to reset the simulation and then press the go button Yow re full of energy and ideas And that s just what we are looking for UBS 2010 All rights reserved Looking for a career where your ideas could really make a difference UBS s Gra
60. E The purpose of this model is show how entropy calculations are made given a probability distribution HOW IT WORKS A symbol turtle agent is used to represent the symbols in the distribution Each symbol owns a frequency variable which is its count The entropy is calculated by summing all frequency counts for all symbols first to find out the total count then by calculating the sum of the negative log to the base 2 of the symbol probabilities HOW TO USE IT First select the size of the alphabet using the alphabet size slider Then select the frequency counts and desired colours using the respective sliders for each symbol The sliders for symbols whose number is greater than or equal to the alphabet size will be ignored Then press the setup button followed by the calculate entropy button Download free ebooks at bookboon com 175 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises THE INTERFACE The model s Interface buttons are defined as follows setup This clears the environment and then draws coloured circles to represent the balls in the distribution calculate entropy This calculates the entropy for the distribution The model s Interface sliders are defined as follows alphabet size This sets the number of symbols coloured balls in the distribution The remaining sliders have the following naming convention lt symbol gt lt number gt lt count gt and
61. ECHNOLOGY amp SCIENCE PROGRAMME Are you about to graduate as an engineer or geoscientist Or have you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com 180 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Expanding the search for searcher 10 by hatching new searcher 12 at node n1 Path cost of searcher 12 40 Expanding the search for searcher 10 by hatching new searcher 13 at node n3 Path cost of searcher 13 40 Expanding the search for searcher 10 by hatching new searcher 14 at node n5 Path cost of searcher 14 30 Expanding the search for searcher 13 by hatching new searcher 15 at node n2 Path cost of searcher 15 60 Expanding the search for searcher 13 by hatching new searcher 16 at node n6 Path cost of searcher 16 50 Expanding the search for searcher 16 by hatching new searcher 17 at node nO Path cost of searcher 17 60 Searcher agent 17 has found the goal node nO The output produced for the A Search on Network 1 is as follows Figure 8 3 1 4 Screenshot of the Environment for the Searching for Kevin Bacon 2 model showing how the A Search using the heuristic Manhattan Distance searches the network shown loaded from the Network 1 data file The searcher agent 9 is starting at node 1 Expanding the search for searcher 9 by hatching new searcher 1
62. Exercises Agent Behaviour 7 Communication THINGS TO TRY Try loading all the language models from the different texts specified by the which text chooser While doing this set different values for the order of the language models by setting the max tree depth slider from 1 to 8 Be careful when this value is set high the language models can take some time to load and the visualisation becomes extremely cluttered Try extending the maximum size of the training text using the max text size slider and see what affect this has on the complexity of the language models Try changing the window size slider to see what effect this has on the Most Active Recently in Window monitor EXTENDING THE NETLOGO MODEL Try a variation of the NetLogo model where most of the nodes are hidden and only the ones chosen to be visualised by the pick these streams and the pick these events choosers are shown in the environment This will mean much larger language models can be loaded since the reset layout procedure for the spring layout type is what slows down the program when a language model is loaded An alternative animation could also be tried where only the active part of language model is animated i e Everything in the language model remains hidden until it is activated If entropy and code length calculations are added like in the Cars Guessing Game NetLogo model then the model can be extended to perform with a high degree of accura
63. Firebreak 6 1 2 Solution to 6 1 2 This model is an extension of the Fire model that allows users to add firebreaks extra forest and ignition points A satellite or aerial image can also be imported into the environment Fireflies 6 3 1 Solution to 6 3 1 6 5 1 Solution to 6 5 1 This model simulates the synchronization of flashing behaviour in fireflies In NetLogo s Models en ree gt Fireflies Flocking 6 3 1 Solution to 6 3 1 6 5 1 Solution to K 5 1 This model simulates the De behaviour of Dues In Pa s Models Library Biology gt Flocking Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Preface Flocking Perspective Demo 6 4 2 This is a variation of the Flocking model that offers different perspectives In NetLogo s Models Library Perspective Demos gt Flocking Perspective Demo http ccl northwestern edu netlogo models FlockingPerspectiveDemo Flocking With Obstacles 6 7 2 This model is an extension of the Flocking model with obstacles added http files bookboon com ai Flocking With Obstacles html Follow and Avoid 6 7 3 This model is an attempt to recreate boids see Craig Reynold s work that use seeking and fleeing steering behaviours http files bookboon com ai Follow And A void html Heatbugs 6 5 1 Solution to 6 5 1 This model simulates the behavior of biologically inspired agents tha
64. Hungarian mathematician who was prolific at publishing papers Paul Erd s has been chosen as the focal point for measuring how well connected a person is amongst the mathematical community The degrees of separation measure is based on whether as person has published a paper with Paul Erd s which equates to an Erd s number of 1 or if not then published a paper with someone who has published a paper with Paul Erd s an Erd s 7 Communication Download free ebooks at bookboon com 52 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication number of 2 or if not then published a paper with someone who has published a paper with someone who has published a paper with Paul Erd s an Erd s number of 3 and so on The model allows the user to select either Paul Erd s or Kevin Bacon as the centre of the universe Then it can measure the average degree of separation using Dijkstra s algorithm or it can search for the node associated with these persons using various distributed algorithms for sharing of knowledge between agents such as word of mouth WHAT IS ITS PURPOSE The purpose of this model is to demonstrate some important concepts in relation to communication amongst agents in a network such as the small world phenomenon degrees of separation super nodes in peer to peer networks and communication via word of mouth In Paris or Online International programs taugh
65. IT This model shows how a language model can be constructed from some training text and then used to predict text i e play the Shannon Guessing Game a game where the agent human or computer tries to predict upcoming text one letter at a time based on the prior text it has already seen Note There is some possible confusion of the term model for this NetLogo model NetLogo uses the term model to refer to a program simulation This can be confused with the use of the term model in the phrase language model which is a mechanism for assigning a probability to a sequence of symbols by means of a probability distribution In the information listed below the two types of models will be explicitly distinguished by using the phrases NetLogo model and language model The type of language model used in this NetLogo model is a fixed order Markov model that is according to the Markov property the probability distribution for the next step and future steps depends only on the current state of the system and does not take into consideration the system s previous state The language model is also fixed order that is the probability is conditioned on a fixed length prior context HOW IT WORKS The language model is based on the PPMC compression algorithm devised by John Cleary and Jan Witten It uses a back off mechanism called escaping to smooth the probability estimations with lower order
66. If they do they will overpower the missionaries and then eat them The problem is the following Is it possible to get all the missionaries and cannibals safely across to the other side of the river To solve this problem search needs to be employed to try out the Download free ebooks at bookboon com 91 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search different scenarios This model implements some of the classic search algorithms and these can be applied to this toy problem to see how the different search strategies perform against each other WHAT IS ITS PURPOSE The purpose of this model is to show how some of the classic search algorithms such as breadth first search and depth first search can be implemented in NetLogo and also how they can then be applied to solving a classic toy problem in order to compare how they perform HOW IT WORKS The model implements the search algorithms in a novel way by making use of an agent oriented approach which is relatively easy to implement in NetLogo Rather than use a queue data structure to implement the classic search algorithms this is a standard solution see for example Russell amp Norvig s AI textbook the model instead adopts a purely agent oriented solution where the information is distributed amongst NetLogo agents These agents conduct the search by transferring information to other agents that continue the search Hence an explic
67. S For other examples of the use of event maps for event stream processing see the Language Modelling Central Park Events and Wall Following Events NetLogo models Exercise 9 8 3 Have a play with the Knowledge Representation NetLogo model to try out the different methods of representation To select the problem use the select problem chooser in the Interface Select the different types of knowledge representation by selecting the appropriate option using the KR type chooser Try different settings to see what happens for example you can use the change layout button combined with the sliders on the right to make the displayed network more visually appealing Also try out the go reason button to see what happens Explain what you observe Exercise 9 8 4 Try adding the knowledge for your own problem into the Knowledge Representation model described in Exercise 9 8 2 Observe while you are doing this the problems you have with formulating the knowledge for each of the different types of knowledge representation What are the advantages or disadvantages of using the different types How could they possibly be improved given the insights you have gained by the knowledge representation process Is there yet another new type of knowledge representation that would be better suited for the task Exercise 9 8 4 The three example problems tackled by the Knowledge Representation NetLogo model are toy problems rather than real world proble
68. a roast meal e the knowledge of how to drive from Bangor to Betwys y Coed in North Wales e the knowledge of when you last went to a wedding or a funeral e the knowledge for a robot of how to lift its robotic arm e the knowledge that there are choices to be made when you are at a junction in a maze e the knowledge that if a Takehe lays an egg then it must be a bird e the knowledge that we cannot be in two places at the same time Download free ebooks at bookboon com 118 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge e the knowledge required to solve a mathematical equation e the knowledge of which wild mushrooms are poisonous e the knowledge that the mother of Jesus is Mary e the knowledge of how to open a tin of spaghetti e the knowledge of what a Pohutukawa tree looks like e the knowledge of what bagpipes sound like e the knowledge of what haggis tastes like e the knowledge of how it feels to stroke a cat e the knowledge of what your face looks like e the knowledge that Elvis is still alive e the knowledge of what strong coffee smells like when hung over Pohutukawa Tree e the knowledge of how to pronounce the word Llangollen correctly e the knowledge of when someone is being polite or being rude e the knowledge of what might happen when you do something wrong e the knowledge of how to bear oneself when in the presence of Her Majesty the Queen e the knowledge requ
69. age model is then used to process a test text sequence Each stream event from the training data is represented by a state turtle agent and paths between states are represented by path link agents When the language model is used for prediction such as when playing the Shannon Guessing Game as in this NetLogo model then walker turtle agents are used to walk the network of states to determine which of the current contexts are active These are then used to make the probability estimates Text turtle agents are used for maintaining information about a particular text such as sums and a variable for storing the language model HOW TO USE IT To use this NetLogo model perform the following steps 1 Select a training text to load using the which training text slider 2 Load and build a language model for it using the load training text button 3 4 Then predict the testing text one character at a time by pressing the predict text button Select a testing text to perform the prediction on using the which testing text slider THE INTERFACE The model s Interface buttons are defined as follows load training text This loads the training text specified by the which training text slider At the same time it constructs the language model from it load testing text This loads the testing text specified by the which testing text slider predict text This predicts the text in the testing text one character at a time
70. ally building links to other random nodes Star central hub This creates a network with one node the Kevin Bacon node i e the goal node as the central hub and all other nodes linked to it and not to any other node Hierarchical This creates a tree network with the Kevin Bacon goal node at the root layout type This specifies how the network should be laid out when it is visualised no of nodes This is the number of nodes to place in the network no of super nodes This is the number of super nodes to place in the network These are nodes that usually will have significantly more links than standard nodes as specified by links per super node slider links per node This specifies the maximum number of links a standard node will have The actual number chosen for a particular node will be a random number between and this number links per super node This specifies the maximum number of links a super node will have The actual number chosen for a particular super node will be a random number between 1 and this number search behaviour This specifies what search strategy the agents should employ when performing the search A full description of the different search strategies can be found in Chapter 8 of the book Artificial Intelligence Agent Behaviour P start node number This is the node where the searcher agent is placed when it starts the search Specifying 0 will mean that the searcher agent will be placed
71. ange layou y move layout layout type lecision t v New Zeala 7 J spring v ml C Tor ticks 0 30 Output spring constant 0 10 Matching tree state with stream can fly event Yes bird Ka Matched tree state with stream can fly event Yes Matching tree state with stream parrot event Yes Matched tree state with stream parrot event Yes alpine Nd f Matching tree state with stream alpine event Yes bird Ke ar can fly Yi sport white throat A User One Of Is the bird an apline bird gt Yes s i has long beak Ye EENE has long beak Figure 9 8 2 Screenshot of the Interface for the Knowledge Representation model after the setup knowledge button has been pressed followed by the go reason button The type of knowledge representation has been set to decision tree using the KR type chooser and the problem has been set to the New Zealand birds problem The Output box shows a trace of the initial stages of the reasoning process WHAT IS IT This model shows how the standard types of knowledge representation such as logic frames rules and decision trees can be implemented in NetLogo The model also shows how to perform reasoning using the different types of knowledge representation A new type of knowledge representation called event maps is also implemented Event maps are related to the concept of Event Stream Processing ESP where events
72. arch behaviours to see which one is the most effective Find out how well each search algorithm performs against the four evaluation criteria time complexity space complexity optimality and complexity Relate the theory to what happens in practice atter what you want out of your future ber an_employer with a broad range of operations in a load of countries will always be the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our web site www volvogroup com e look rward to getting to know you VOLVO AB Volvo publ WWW YONOGTOUD OOM Vowo Trucks Rexauer Taucks Macx Toucxks Vowo Buses Vowo Coxsteucrion Eoupwent Vowo Pewa Vowo Aero Vowo IT Vowo Financat Sernces Vowo3P Vowo Powenrmam Vowo Parts Vowo Tecumooer VowoLoasncs Business Anca Asn Download free ebooks at bookboon com 95 Artificial Intelligence Exercises Agent Behaviour 8 Search Try figuring out how many solutions there are to the problem the one shown in the animation is not the only one Try out the different heuristics for the informed searches Greedy Best First Search and A Search Do they have any effect on how effective the search is EXTENDING THE MODEL Try adding other search algorithms or adding your own variations to the ones implemented in the mode
73. arches the network shown loaded from the Network 1 data file The searcher agent 9 is starting at node 1 Expanding the search for searcher 9 by hatching new searcher 10 at node n2 Expanding the search for searcher 9 by hatching new searcher 11 at node n4 Expanding the search for searcher 11 by hatching new searcher 12 at node n1 Expanding the search for searcher 11 by hatching new searcher 13 at node n5 Expanding the search for searcher 11 by hatching new searcher 14 at node n7 Expanding the search for searcher 10 by hatching new searcher 15 at node n1 Expanding the search for searcher 10 by hatching new searcher 16 at node n3 Expanding the search for searcher 10 by hatching new searcher 17 at node n5 Expanding the search for searcher 16 by hatching new searcher 18 at node n2 Expanding the search for searcher 16 by hatching new searcher 19 at node n6 Expanding the search for searcher 13 by hatching new searcher 20 at node n2 Expanding the search for searcher 13 by hatching new searcher 21 at node n4 Expanding the search for searcher 13 by hatching new searcher 22 at node n6 Expanding the search for searcher 13 by hatching new searcher 23 at node n8 Expanding the search for searcher 12 by hatching new searcher 24 at node n2 Expanding the search for searcher 12 by hatching new searcher 25 at node n4 Expanding the search for searcher 15 by hatching new searcher 26 at node n2 Expanding the search for searcher 15 by hatching new searcher 27
74. arching Mazes model For an insight into the Manhattan distance and Euclidean distance metrics used as heuristics for the Informed searches see the Manhattan Distance model Download free ebooks at bookboon com 90 Artificial Intelligence Exercises Agent Behaviour 8 Search Exercise 8 2 2 Missionaries and Cannibals NetLogo Model Try out the Missionaries and Cannibals model in NetLogo M4 Sticks 14 go once go forever go animation BSSSSSSSSHRA EES RAANANNNNOAOAAAAG SOPERNWWOOHENNWWOOHENNWWOORENNW Active Searcher Agents 3 Maximum Active Searcher Agen 3 Total Searcher Agents 16 Figure 8 2 2 Screenshot of the Interface for the Missionaries and Cannibals model for after the setup button has been pressed followed by the go forever button and then the go animation button with the slider values as shown in the image WHAT IS IT This model applies standard search algorithms to the classic search problem called Missionaries and Cannibals In this toy problem there are 3 missionaries and 3 cannibals They have arrived together on one side of a river and they all wish to cross to the other side Luckily there is available a boat for crossing the river However there are two catches to the problem The first catch is that only two people can fit in the boat at any one time The second catch is that the cannibals are not allowed to outnumber the missionaries at any stage
75. astest making the turtles more likely to bump into one another The faster turtles try to get past the slower ones Download free ebooks at bookboon com 25 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour HOW TO USE IT Decide on the size of the population then press the setup button then press the go button The slider bars allow various parameters to be altered To draw the boid trails press the trail button to clear them press the clear button To follow one particular boid press the follow button THE INTERFACE The model s Interface buttons are defined as follows Setup This sets up the environment with a black path across the middle and the rest shown as a green texture to represent the surrounding grass A number of boids determined by the population variable are created and placed at a random location within the black path Go The boids start moving along the path and after a while most of them end up heading in the same direction Follow The simulation follows the path of one particular boid The financial industry needs a strong software platform That s why we need you Find your next challenge at www simcorp com careers www simcorp com m S l m Co rp MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 26 Artificial Intelligence Exercises Agent Behaviour Trail This draws the paths made by
76. at bookboon com 69 Artificial Intelligence Exercises Agent Behaviour 7 Communication THINGS TO NOTICE Notice how well the adaptive distribution does compared to the source distribution The source distribution will have the lowest code length total compared to the others but usually the nearest distribution will be the adaptive one unless the counts for one or more of the Agent s distribution exactly match the source distribution counts If the source distribution has equal counts then the adaptive counts will rise at a similar rate and as a consequence the red line in the plot will rise diagonally from bottom left to top right Why A horizontal line in the plot indicates a fixed distribution Modifying the distribution counts in the middle of the animation will result in the lines in the plot changing to reflect this THINGS TO TRY Try changing the distribution counts in the sliders to see what affect this has on the entropy and code length calculations and on what happens in the plot Note that the lower colours blues and pinks will cause the greatest shifts in the line plots Why Can you achieve a situation where the code length of one of the three non adaptive Agents is better than the adaptive Agent i e the code length is smaller and closer to the source agent s code length total 7 9 Language Modelling Exercise 7 9 1 Language Modelling NetLogo Model Try out the Language Modelling model in NetL
77. ath has been designated as complete using the end path button or an intervening object has been drawn which ends the path by default atter what you want out of your future ber an_employer with a broad range of operations in a load of countries will always be the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our web site www volvogroup com e look rward to getting to know you VOLVO AB Volvo publ WWW YONOGTOUD OOM Vowo Teucxs Rexauer Trucns Macx Toucxs Vowo Buses Vowo Coxsteucrion Eoupwewt Vowo Pewa Vowo Aero Vowo IT Vowo Financat Sernces Vowo3P Vowo Powenrmam Vowo Parts Vowo Tecumooer VowoLoasncs Business Anca Asn Download free ebooks at bookboon com 129 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge draw object This will draw an object at the location specified by the next mouse click The object type and colour is specified by the obj ect t ype and object colour choosers and its size by the obj ect size slider erase object This will erase the nearest object if there is one Make sure to click on the centre of the object otherwise it will not be erased select object This will allow the user to drag a selected object using the mouse in order to change its position The selected object i
78. aughter of Ruby 10 Some cats are small 11 All cats eat meat 12 Cats have whiskers and a tail SANAMAWN gt Represent the knowledge contained in each of the sentences above using the two knowledge representation paradigms production rules and first order logic Which sentences do you have problems with when trying to represent the knowledge and why Exercise 9 9 2 Convert the following predicate logic sentences into English sentences i Ax Car x A Red x ii Vx Cake x gt Sweet x iii AxVy Person x A Cake y AaLikes x y Download free ebooks at bookboon com 143 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Exercise 9 9 3 Convert the following sentences into predicate logic sentences i All cats like fish ii Bananas are either green or yellow or brown iii Some people eat everything 9 10 Reasoning using rules and logic Exercise 9 10 1 Consider the following rule based expert system ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on making the world a better place for people business and society Join us In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To a
79. ave you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com 47 Artificial Intelligence Exercises Agent Behaviour 7 Communication 7 3 Communication via communities of agents Exercise 7 3 1 Language Change NetLogo Model Try out the Language Change model in NetLogo threshold val and sink state only apply for individual and threshold updating algorithms when logistic is off there s a built in bias towards grammar 1 alpha only applies for the reward updating algorithm when logistic is on Figure 7 3 1 Screenshot of the Interface for the Language Change model after the setup everything button has been pressed followed by the layout button to unclutter the network and then the go bution with the slider and switch values as shown in the image Read the What is it and Things to Notice sections in the Information for the model Then try out the exercises suggested in the Things to try and Extending the Model sections In particular the Information in the model states that language users tend to arrive at using just one grammar but in some cases they may not all converge to the same grammar Try to find out what conditions cause the language users to converge to a single grammar and what conditions result in multiple grammars being used Download free ebooks at bo
80. bookboon com 41 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour vradius angle This defines the radius angle of the boid s vision cone radius length This defines the radius length of the boid s vision cone HOW TO USE IT Press the Setup button first then press Go To see where the boid wanders press Plot You can draw extra obstacles by pressing the Draw Obstacle button and then holding down the mouse at the point where you want the obstacles to be drawn You can change the frame of reference so that the visualisation is centred around where the boid currently is situated by pressing the Follow Wanderer button THINGS TO NOTICE Setting the boid speed to 0 1 rate of random turn to 40 radius angle to 300 radius length to 1 and pressing the Plot button once followed by moving the speed slider just below the Information tab in the Interface from normal speed to faster will result in the boid rapidly covering the entire environment while reliably avoiding the blue obstacles Increasing the radius length to 5 while keeping the other variables the same discernibly changes the behaviour of the boid Instead of covering most of the environment the boid covers a rectangular path of width 4 to 5 around the outside of the environment but indented by about 3 to 4 patches in from the outer boundary The boid seems to refrain from going inside of the rectangular path Sometimes the boid can get stuck spinning
81. by pressing the Settings button in the Interface Download free ebooks at bookboon com 130 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge The model s Interface choosers sliders and Input box are defined as follows path colour This sets the colour that is used to draw the path when the draw path button is pressed path gsize This sets the pen size that is used to draw the path when the draw path button is pressed object type This sets the type of object that is drawn when the draw object button is pressed object colour This sets the colour of the object that is drawn when the draw object button is pressed object size This sets the size of the object that is drawn when the draw object button is pressed background colour This sets the colour of the background that is drawn when the clear background button is pressed patches path size This sets the size of the path that is drawn when the draw patches path button is pressed turtles path size This sets the size of the path that is drawn when the draw turtles path button is pressed text This sets the text that is drawn when the draw text button is pressed text colour This sets the colour of the text that is drawn when the draw text button is pressed THINGS TO TRY Try drawing your own map perhaps of your house apartment or flat or of the local school Note while you are making the map the problems t
82. cher 29 at node n19 Height of searcher 29 30 Expanding the search for searcher 27 by hatching new searcher 30 at node n8 Height of searcher 30 10 Expanding the search for searcher 27 by hatching new searcher 31 at node n12 Height of searcher 31 25 Expanding the search for searcher 27 by hatching new searcher 32 at node n14 Height of searcher 32 25 Expanding the search for searcher 27 by hatching new searcher 33 at node n18 Height of searcher 33 20 Expanding the search for searcher 30 by hatching new searcher 34 at node n3 Height of searcher 34 5 Expanding the search for searcher 30 by hatching new searcher 35 at node n7 Height of searcher 35 20 Expanding the search for searcher 30 by hatching new searcher 36 at node n9 Height of searcher 36 20 Expanding the search for searcher 30 by hatching new searcher 37 at node n13 Height of searcher 37 15 Expanding the search for searcher 34 by hatching new searcher 38 at node nO Height of searcher 38 0 Searcher agent 38 has found the goal node n0 Download free ebooks at bookboon com 183 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 9 1 2 What is a Knowledge Base and what must it be made of There are many answers to these questions in the literature For example one possible answer might be as follows A Collection of concepts about a domain or domains But what is a conc
83. cific point press the ignite forest button and click the mouse button at the required location To have the fire start at the left of the environment as with the Fire model then press the ignite on the left button To create more forest press the draw forest button To create a firebreak press the draw firebreak button Firebreaks can be removed using the remove firebreaks button Burnt forest can be unburnt using the unburn forest button This means that a simulation for the same forest can be run multiple times using different firebreaks and ignition points THE INTERFACE The Interface buttons are defined as follows setup This will reset the simulation and create a random forest go This will start the simulation If there have been any fire ignition points placed in the environment then the fire will start spreading to adjacent trees draw forest This will draw more trees where the mouse is subsequently clicked import forest This will import an environment based on a small area in a satellite image taken of Chateaubriant France unburn forest This will restore the trees to the way the forest was before any trees were burnt Note that this will not work when the satellite image has been imported To restore that forest simply press the import forest button again burn all the forest This will eventually burn all the forest It repeatedly chooses unburnt trees at random each tick to spontaneously combust
84. cise 6 7 7 For the Follow and Avoid model described in Exercise 6 7 3 the behaviour of the three breeds of agents is defined by the wanderers followers and avoiders procedures Explain how these three procedures work Which implements boid like behaviour and which doesn t Exercise 6 7 8 For the Flocking with Obstacles model described in Exercise 6 7 2 explain how the code works to produce flocking In particular explain the behaviour that is being modelled by the following procedures in the model e separate e align e cohere Also detail any emergent phenomena that you witness an obvious example is the flocking but there are others when various variables are changed using the Interface sliders Download free ebooks at bookboon com 46 Artificial Intelligence Exercises Agent Behaviour 7 Communication Please click the advert 7 Communication 7 1 Communication Information and Language Exercise 7 1 1 Compare the different meanings of the following terms by finding out how they are being defined in the literature e communication e information e language 7 2 The diversity of human language Exercise 7 2 1 How many human languages are being used in the world today Which continent has the greatest number of languages in use WHAT S MISSING IN THIS EQUATION MAERSK INTERNATIONAL TECHNOLOGY amp SCIENCE PROGRAMME Are you about to graduate as an engineer or geoscientist Or h
85. cle size This sets the size of the circle cCircle thickness This sets the thickness of the circle text to insert This sets the text that is drawn when the draw text is next pressed The text is right justified rather than left justified i e The end of the text not the start will end up at the point where the mouse is clicked The size of the text can be specified globally using the Settings button at the top of the Interface text colour This sets the colour of the text that is drawn when the draw text button is next pressed The colour is specified in a range from 0 to 140 using NetLogo s colour numbering scheme Sample colour numbers are shown in the small text note at the left bottom of the Interface mark colour This sets the colour of the mark i e a letter x that is drawn when the draw mark button is next pressed The colour is specified in a range from 0 to 140 using NetLogo s colour numbering scheme Sample colour numbers are shown in the small text note at the left bottom of the Interface mark size This sets the size of the mark THINGS TO NOTICE Notice that some of the images do not fit the environment exactly there are white unfilled areas above and below the image in the environment This is because the import drawing command maintains the same aspect ratio of the image i e the width and height have the same ratio relative to each other as the image is scaled The image however is sized to fit
86. cy the task of language identification authorship identification or text categorisation simply by choosing the class i e language author or category of the testing text to be the class of the text used to train the language model which produces the smallest code length on the testing sequence RELATED MODELS See the Central Park Events and Wall Following Events NetLogo models For prediction using these language models see the Shannon Guessing Game NetLogo model 7 10 Entropy of a Language Exercise 7 10 1 Shannon Guessing Game NetLogo Model Try out the Shannon Guessing Game model in NetLogo Download free ebooks at bookboon com 76 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication load training text Text so far The family load testing text Predictions at various depths in the training model s tree Each character is listed followed by its frequency of occurrence and estimated probability depth 2 types 1 tokens 1 escape probability 1 2 J G 2 depth 1 types 1 tokens 3 escape probability 1 4 B G A depth types 28 tokens 370 escape probability 28 398 69 69 398 e 33 33 398 o 27 27 398 i 26 26 398 t 25 25 398 23 23 398 s 22 22 398 r 18 18 398 h 18 18 398 a 16 16 398 13 13 398 1 11 11 398 u 11 11 398 d 10 10 398 g 9 9 398 7
87. d might have one time been joined and importantly to show how the process of continental drift has occurred over millions of years One salient insight that can be gained from the knowledge of how this process has occurred is that species will have to adapt to epoch level changes if they are to survive for a very long time Download free ebooks at bookboon com 166 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises HOW IT WORKS The model uses the import pcolors command to load a map of Australia and New Zealand and convert the image into patch agents In order to simplify the shifting of New Zealand back to Australia these patch agents are replaced with turtle agents yellow and green for Australia and black for New Zealand to signify the traditional colours of the respective countries Then a simple translation on a specific heading with some warping is used to move all the black agents back closer to the yellow and green ones Clearly these simulated transformations are nothing like the type of transformations that have occurred during the real process of continental drift between Australia and New Zealand For example in the model Australia remains static and does not drift in the simulation only New Zealand does However as stated above much more code would be required to get a decent simulation that more accurately represents what occurred during the real life process And this was no
88. defined as follows radius detection This adjusts the vision of the boids boid speed This adjusts the forward step value of each boid per iteration speed factor This reduces the speed of both the Follower and Avoider agents boid random heading This adds a bit of randomness to the boids movements number of agents number of wanderers number of followers number of avoiders These are the number of agents in each respective category Download free ebooks at bookboon com 34 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THINGS TO NOTICE Adjusting the radius bar alters the distance at which both Followers and Avoiders can see This alters the way they behave such as the distance that the Avoider stays away from Wanderers The Avoider sometimes appears to move towards a Wanderer this is an affect resulting from an environment that wraps around THINGS TO TRY Try changing the number of Wanderers Followers and Avoiders Try adjusting the scale bar to see what happens to the Follower Try changing the shape of the Avoider to be rotatable if they are not already this highlights the affects of the wrap around problem EXTENDING THE MODEL The model could be extended to add gradual acceleration and deceleration This would enhance the boids simulation Fast track your career Masters in Management Stand out from the crowd Designed for graduat
89. djacent trees and has no ability to jump over an unwooded area patch The fire is implemented using fire turtle agents The fire spreads in a breadth first manner similar to the breadth first search implemented in the Searching Mazes Missionaries and Cannibals and Searching for Kevin Bacon models Once a fire agent has ignited its nearest neighbours if there are any then it will change its breed to an ember agent whose colour slowly fades from red to black according to the value of the fire fade amount slider and then it will die SimCorp is a leading provider of software solutions for the financial industry We work together to reach a common goal to help our clients succeed by providing a strong scalable IT platform that enables growth while mitigating risk and reducing cost At SimCorp we value commitment and enable you to make the most of your ambitions and potential Are you among the best qualified in finance economics IT or mathematics m www simcorp com S l m Co rp MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 159 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises HOW TO USE IT First press the setup button This will create a forest of randomly spaced trees with density as specified by the density slider Alternatively to import the satellite image of Chateaubriant France press the import forest button To start a fire at a spe
90. ds Searching Mazes 8 2 1 This model applies standard search algorithms to the problem of searching mazes Shannon Guessing Game 7 10 1 This model shows how a language model can be constructed from some training text and then used to predict text i e play the Shannon Guessing Game a game where the agent human or a hana tries to eas BDO text one aa at a time i on the prior text it has already seen http files bookboon com ai Shannon Guessin Download free ebooks at bookboon com 11 Artificial Intelligence Exercises Agent Behaviour Preface Download free ebooks at bookboon com 12 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour 6 Behaviour 6 1 What is behaviour Exercise 6 1 1 Try out the different behaviours for the turtle agent in the Mazes 2 NetLogo model R MS ticks 1403 Empty Maze parameters go forever Hampton Court Chevening House and Butterfly Mazes parameters i The following sliders control the movement for the Random Forward 3 and Random Forward 4 behaviours Mii Figure 6 1 1 Screenshot of the Interface for the Mazes 2 model after the setup button has been pressed followed by the go forever button with the chooser and slider values as shown in the image and with the simulation left running for a short while Do this by changing the value of the turtl
91. duate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates www ubs com graduates a UB S Download free ebooks at bookboon com 171 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises THE INTERFACE The buttons in the Interface are defined as follows setup This resets the simulation so that the wood chips and termites are placed randomly go This runs the simulation The sliders chooser monitor and plot in the Interface are defined as follows number This sets the number of termites that will be created at the start of the simulation density This sets the percentage of patches that will be created as wood chips at the start of the simulation pileness measure These are the functions that can be used to measure the self organisation of the system i e how much of the wood chips have been organised into piles There are two possible functions Number of crowded spaces This counts the number of patches which have more than 7 yellow coloured neighbours Number of empty spaces This counts the number of patches which have more than 7 black coloured neighbours pileness This reports the value of the pileness measure as computed by th
92. dumps into the Output box the knowledge that was loaded by the setup knowledge button reset reasoning This resets the reasoning to start over again from the start state This state is determined by the type of reasoning being performed i e for backward reasoning using tules the start state is the goal state for forward reasoning the start state is the state associated with the first rule in the rules database go reasoning This starts the reasoning process If rules based reasoning is being used the user will be asked whether they wish to perform forward or backward reasoning clear output This clears the Output box change layout This changes the layout if the Layout type chooser is set to spring This can be useful to remove some of the clutter In order to change the layout configuration the Download free ebooks at bookboon com 139 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge user can alter the spring length spring constant and repulsion constant sliders while the change layout button is pressed move layout Sometimes the labels of states in the layout appear off the edge of the environment This button allows the layout to be moved left right up or down by a user specified distance The model s Interface sliders switches choosers and output box are defined as follows KR type This selects the type of knowledge representation e g rules frames decision trees and sema
93. e This can then subsequently be read by agents who visit the node latter on Combined 1 and Combined 2 These combine the knowledge of the word of mouth and blackboard approaches in slightly different ways to create two new hybrid knowledge sharing schemes The first scheme checks the blackboard first then defaults to the word of mouth scheme if no knowledge exists in the blackboard The second scheme does the reverse it will apply word of mouth knowledge first before defaulting to the blackboard knowledge network type This selects the type of network that is created when the setup network button is pressed The type of networks are as follows P2P no super nodes This simulates a peer to peer network with no super nodes P2P has super nodes This simulates a peer to peer network with super nodes P2P random single link This simulates a peer to peer network where each node has only one link with another random node P2P incremental This creates the simulated peer to peer network by incrementally building the links to other random nodes one at a time P2P incremental 1 This creates the simulated peer to peer network by incrementally building links to other random nodes Star central hub This creates a network with one node the Kevin Bacon node i e the goal node as the central hub and all other nodes linked to it and not to any other node Hierarchical This creates a tree network with the Kevin Bacon goal node at the ro
94. e a Rogerian psychotherapist but Harry does try to be a bit paranoid like Parry HOW IT WORKS The model makes use of an extension to NetLogo called re for regular expressions The extension allows developers in NetLogo to specify and match regular expressions In the NetLogo code a setup rule command is used to specify the regular expression and a list of possible chatbot responses If the chatbot matches the user s reply against the regular expression in the rule then it will choose one of the replies from the response list 149 Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour 10 Intelligence The extension re adds the following commands to NetLogo compile pattern This compiles the regular expression pattern passed as the first argument set target This sets the target string that you want to search for to that specified by the first argument run matcher This tries to match the regular expression in the target string get group length This returns the number of groups get group This returns the item in the group at the position specified by the first argument get starts length This returns the length of starts get starts This returns the item in starts at the position specified by the first argument get ends length This returns the length of ends get ends This returns the item in ends at the position specified by the first a
95. e behaviour chooser in the Interface Characterise each behaviour by the set of actions the turtle carries out when performing that behaviour Characterise these actions that comprise each behaviour by the set of movements that the turtle performs when carrying out the action Which behaviour seems to be the most effective on a particular maze and which seems to be most effective across all mazes Which behaviours end up with the turtle agent getting stuck in an endless loop in some of the mazes Download free ebooks at bookboon com 13 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Exercise 6 1 2 Try out the Fire model in NetLogo How does the fire behave when the value of the density slider is increased or decreased Characterise the behaviour of the fire by the patterns of behaviour that result from the different slider values for example the amount of the forest that is burnt the pattern of burning and the pattern of the remaining unburnt trees Extend the Fire model by adding a draw firebreak button to allow the user to create firebreaks in the forest to stop the spread of the fire Add two further buttons draw forest and ignite forest for allowing the user to draw extra trees in the forest and to specify specific locations for where the fire starts rather than having it start in a line on the left of the environment as with the Fire model For example in the image below the firebreak has been d
96. e can try to find solutions that change the problem in some way or we can try to change the physical properties of the universe i e have it work in a virtual environment but it would not work when transferred to a real environment or we could create an illusion i e it would look like the problem was solved but like magic it would just be trickery A secondary purpose is to highlight the process of problem solving that humans are very good at No computer program is capable of creating the variety of solutions that are shown in this model For that matter no computer programs are yet capable of understanding the problem itself In Paris or Online International programs taught by professors and professionals from all over the world BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizonsuniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 153 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence HOW IT WORKS The model implements various types of agents to try to simulate water flow turtle agents patch agents agents that are boids from Craig Reynolds
97. e function chosen by the pileness measure chooser pileness plot This plots the pileness value for each tick of the simulation THINGS TO TRY Try devising your own pileness measures or alternative functions to measure the self organisation of the system RELATED MODELS See the Termites model in the Models Library Download free ebooks at bookboon com 172 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 6 3 4 Self organisation itself is an emergent property of the system so your answer to Exercise 6 3 3 provides one way to measure when an emergent property of this type has occurred But the measure is problem specific i e the pileness measures devised for Exercise 6 3 2 will work for the Termites and State Machine Example 2 models but will be less useful for other models A more difficult problem is whether any emergent property or any form of self organisation can be determined when it occurs This might be useful when examining many different simulations to determine if some interesting or unforeseen property of the system arises In some multi agent systems however this might happen rarely Computers are good at running many simulations and at collecting statistics about them but when a property of the system cannot be predicted at startup due to its complex nature then it can become problematical to signal that an emergent property has arisen or that s
98. e path behaviour This defines the boid s behaviour It can be set to the following values Basic crowd following This is very basic crowd path following behaviour without collision avoidance With collision avoidance This combines crowd path following behaviour with collision avoidance if the collision slider is set to On Download free ebooks at bookboon com 27 6 Behaviour Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THINGS TO NOTICE Notice how the turtles come close to the sides but manage to avoid them How is this achieved and what is the mechanism for keeping them in track Notice how the speed variances are produced Each turtle has a unique speed setting Notice that the colours are randomised based on the chromatic point at which it was created This allows a random looking range of colours to be produced while avoiding very dark and black colours which would clash with the path Notice the Follow turtle button and how a single turtle can be analysed more closely the 3D button also is of interest here Notice that the turtles only appear on the black patches How is this done Notice that when the boid speed variable is set to a large number 5 0 say some of the boids start moving off the path and end up going across the surrounding green grass Why is this Is there some way we can code the turtle agent s movement to prevent this from happening THINGS TO TRY Try alt
99. e search each step It is only used by the Multi Agent Depth First Search allow revisited states This is a flag that if set to On will allow the search to proceed to already visited states The model s Interface monitors are defined as follows Active Searcher Agents This shows the current number of active searcher agents still participating in the search Maximum Active Searcher Agents This shows the maximum value that the Active Searcher Agents monitor has achieved throughout the search Total Searcher Agents This shows the total number of searcher agents that the search has used Download free ebooks at bookboon com 94 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search IDS Depth This is the current depth during the execution of the Iterative Deepening Search THINGS TO NOTICE Notice the effect of turning the flag allow revisited states On and Off Why is turning it Of f so effective at dramatically reducing the search for the different search behaviours Notice that some of the searches if repeated with the same settings end up in different places for example try Greedy Best First Search several times using the heuristic Min no boat trips needed and with the flag allow revisited states set to Off Why is this THINGS TO TRY Which search behaviour seems to be the best at this problem Which seems to be the worst Try out each of the different se
100. ed by the agents repeatedly moving 10 steps in three directions forward left and right How the grid is built up is best seen by continually pressing the go once button in the Interface at the start of each search TT Le TUEA TT Figure 8 5 2 Screenshots of different movement behaviours for the Searching Mazes model Download free ebooks at bookboon com 109 Artificial Intelligence Exercises Agent Behaviour 8 Search For this exercise try reproducing the turtle movement behaviour for the Empty Maze as illustrated in Figure 8 5 2 Then try creating your own patterns by changing the values of the choosers and sliders 8 6 Informed search Exercise 8 6 1 Manhattan Distance Model Try out the Manhattan Distance Model in NetLogo Manhattan Distance http files bookboon com ai Manhattan Distance html mds ticks 0 3D Sin M4 lt gt v c 30 plot euclidean distance Euclidean Distance 25 4558 plot manhattan distance Manhattan Distance 36 Figure 8 6 1 Screenshot of the Interface for the Manhattan Distance model for after the Setup button has been pressed followed by the plot euclidean distance button and then the plot manhattan distance button four times in succession WHAT IS IT This model shows how Manhattan distance is calculated This is compared with Euclidean distance in the model Download free ebooks at bookboon com 110 Please click the adv
101. ed costs Each searcher agent expands the search by seeing if the following actions are possible two missionaries get in the row boat and row to the other side two cannibals get in the row boat and row to the other side one missionary and one cannibal get in the row boat and row to the other side just one missionary gets in the row boat and rows to the other side just one cannibal gets in the row boat and rows to the other side Note that the states of the search for this problem can be represented by a 3 tuple missionaries on start side of river cannibals on start side of river boats on start side of river Hence the start state is represented by the tuple 3 3 1 and the goal state by 0 0 0 In the Interface a graph using parallel co ordinates is used to visualise the search as it proceeds An animation of one possible solution to the problem is also provided below the graph The graph is updated as the animated solution proceeds HOW TO USE IT To initialise the search press the setup button in the Interface A person shape will appear at the start state 3 3 1 at the top left of the graph The goal state is at the bottom of the graph so a rough estimate of whether a search is making progress is how far it has progressed downwards To start the search either press the go once button to make the search proceed one step at a time or press the go forever button to make the search proceed continuously unt
102. edge sharing behaviour mid stream to see if there is any noticeable effect For example try switching from a Word of Mouth sharing method to a Blackboard sharing method then back again EXTENDING THE MODEL Try adding other knowledge sharing mechanisms or adding your own variations to the ones implemented in the model Download free ebooks at bookboon com 60 Artificial Intelligence Exercises Agent Behaviour 7 Communication Try adding your own type of network either randomly created or with a correspondence to a network in real life You will need to add input routines to create the network layout for the latter Instead of having the walker agents searching for a single node in the network try adding resources to each node so that the simulation is similar to the resource discovery problem where a resource such as an MP3 file might appear at multiple nodes throughout the network and each node will have many resources up to tens of 1000s You will need to add a resources variable to the nodes own field where the resources is a table that contains the names of the resources that the node holds locally Try turning off the visualisation of the network by using the hide turtle command to hide all agents and links This will allow you to run much larger simulations up to 100 000s of nodes NETLOGO FEATURES Note the use of the hatch and die commands to create new child walkers at each node and terminate all existin
103. eir knowledge by writing what they know into communal blackboards that are found at each node This can then subsequently be read by agents who visit the node latter on Two further knowledge sharing methods combine these two approaches word of mouth and blackboard in slightly different ways to create two new hybrid knowledge sharing schemes A breed of walker turtle agents is used for the agents that move throughout the maze trying to get to the goal These searcher agents maintain information about the current state of the search such as the path they have taken Each searcher agent expands the search by following the outgoing paths that lead out of the current node that were suggested as a result of the method of knowledge sharing they adopt A breed of dwalker agents are used to walk the network during the execution of Dijkstra s algorithm This can be used to calculate the degree of separation for every node in the network from the goal node HOW TO USE IT To initialise the search and create a random network whose type is specified by the network type chooser press the setup network button in the Interface To setup a walker agent to search the network first press the reset blackboards and reset walkers buttons to clear blackboards and remove any existing walkers then press the setup walkers button to create new ones Several walker turtle agents whose number is specified by the new walkers per tick slider will then be
104. el 8 3 Uninformed blind search Exercise 8 3 1 Show how the Breadth First Search and Depth First Search algorithms search the network shown below Show the list of active searcher agents along with the nodes they are visiting The start node is node n1 and the star node is the goal node i e node n0 Expand the nodes in the order of their node number For example expand node n1 before every other node expand node n2 before every other node except node n1 and so on Also allow nodes to be revisited Download free ebooks at bookboon com 106 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search Repeat this exercise for the following uninformed search algorithms Uniform Cost Search Multi agent Depth First Search Depth Limited Search Iterative Deepening Search edu eus Xe It s only an opportunity if you act on it 107 6007 A a swasis yay 2031 IKEA SE STUDENT Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour 8 Search 8 4 Implementing uninformed search in NetLogo Exercise 8 4 1 Compare how the expand path and expand paths procedures are defined for the three different models the Searching Mazes model the Missionaries and Cannibals model and the Searching for Kevin Bacon model How do the differences in the way they are defined reflect the nature of the search problem How do th
105. el from text button and then the go walkers button with the slider values as shown in the image and ideas And that s You re full of energy UBS 2010 All rights reserved just what we are looking for Looking for a career where your ideas could really make a difference UBS s Graduate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates X UBS www ubs com graduates Download free ebooks at bookboon com 71 Artificial Intelligence Exercises Agent Behaviour 7 Communication WHAT IS IT This model shows how a language model can be constructed from some training text The type of language model is a fixed order Markov model that is according to the Markov property the probability distribution for the next step and future steps depends only on the current state of the system and does not take into consideration the system s previous state The language model is also fixed order that is the probability is conditioned on a fixed length prior context Note There is some possible confusion of the term model for this NetLogo model NetLogo uses the term model to refer to a program simulation This can be confused with the use of the term
106. elated problems occur for the neural network connectionist layer it does not deal with the issue of embodiment What is required is some concrete formalisation or representation of the knowledge i e with physical properties related to embodiment In Paris or Online International programs taught by professors and professionals from all over the world BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizonsuniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 186 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises For example with conceptual spaces the concept of colour is represented by three dimensions and taste by four dimensions This gives a direct correspondence to physical properties which are based on the lower connectionist layers where the signals are being processed of brightness for colour for example At the top level concepts can be structured to form symbolic representations For the other three senses a full answer should look at some possible quality dimensions see G rdenfors conceptual spaces theory that describe
107. elf organisation has occurred especially if the observer of the simulations usually us has no knowledge in advance that they could occur Solution to Exercise 6 5 1 The Moths and Virus models do not exhibit any self organisation so therefore are not examples of swarm intelligence The other models all make use of the environment in some manner through stigmergic local knowledge to co ordinate their activities to produce complex structures through self organisation For the Ants and Termites models physical objects i e chemical scent and woodchips are placed or moved about in the environment and it is the way that this is done that leads the agents to self organise For the other three models the stigmergic local knowledge is present in the local positions or localised flashing of the other agents Solution to Exercise 7 7 1 For working out solutions to Exercise 7 71 try out the Entropy Calculator model in NetLogo Download free ebooks at bookboon com 173 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises F a p Tapi setup calculate entropy VA sw ticks 0 30 siz Wiz Tr Wiz Wiz symbol 0 probability 1 12 negative log probabilit symbol 1 probability 5 12 negative log probabilit symbol 2 probability 3 12 negative log probabilit symbol 3 probability 1 12 negative log probabilit symbol 4 probability 1 12 negative log probabil
108. ence London Business School s Masters in Management will expand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School Regent s Park London NW1 4SA Applications are now open for entry in September 2011 United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Contents 8 Search 84 8 1 Search Behaviour 84 8 2 Search Problems 84 8 3 Uninformed blind search 106 8 4 Implementing uninformed search in NetLogo 108 8 5 Search as behaviour selection 108 8 6 Informed search 110 8 7 Local search and optimisation 113 8 8 Comparing the search behaviours 115 9 Knowledge 117 9 1 Knowledge and Knowledge based Systems 117 9 2 Knowledge as justified true belief 118 9 3 Different types of knowledge 118 9 4 Some approaches to Knowledge Representation and AI 119 9 5 Knowledge engineering problems 124 9 6 Knowledge without representation 124 9 7 Representing kno
109. enhance the boids simulation NETLOGO FEATURES The code uses the in cone command to simulate the boid s cone of vision RELATED MODELS See the following models Crowd Path Following Flocking With Obstacles Follow and Avoid Obstacle Avoidance 1 Obstacle Avoidance 2 Vision Cone Example 2 Wall Following Example 2 These are basic implementations of various Craig Reynold s steering behaviours for boids Another boid related model is the Biology Flocking model in the Models Library Exercise 6 7 5 Obstacle Avoidance 2 NetLogo Model Download free ebooks at bookboon com 39 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour W4 gt ticks 3872 3D Setup Go a vy 30 NN Boid speed adjusts the step boid sp 0 1 value of each boid per iteration Ss ee rate of random turn 40 degrees radius angle 300 degrees 9 ae ee radius length 1 length Draw Obstacle al Ka x Follow Wanderer Plot Figure 6 7 5 Screenshot of the Interface for the Obstacle Avoidance 2 model after the setup button has been pressed followed by the go and plot buttons WHAT IS IT This model implements a boid see Craig Reynold s work that employs basic obstacle avoidance steering behaviour HOW IT WORKS It does this by generating a wanderer turtle boid that simple wanders around randomly in the environment avoiding the obstacles The boid is implemented using NetLogo s in cone command that imple
110. ense reasoning and these are well known stumbling blocks in Artificial Intelligence A conversational agent also needs to understand the rules of conversation as it has to have a conversation with humans e g speech and dialogue acts in A I are possible solutions but only go a step of the way towards producing an agent oriented system with the ability to hold a genuine conversation with a human Download free ebooks at bookboon com 202
111. ents that are added each tick of the simulation time to live This sets how long the walker agents should remain active if they have not found the goal node before being killed off new walkers per node This sets the number of walkers that are created to continue the search in multiple different directions when each walker visits each node Setting this number higher than 1 will quickly flood the network in most circumstances percent in both When the network is created a number of nodes will be created that are connected to the Paul Erdos node i e a path exists that will eventually lead to Paul Erdos and a number will be created that are connected to Kevin Bacon This slider controls how many of the nodes will be doubly connected i e paths exist from the node that will lead to both the Paul Erdos and Kevin Bacon nodes The model s Interface plots monitors shown on the right are defined as follows Number of walkers This plot shows the number of walkers per tick Successful searches This plot shows the number of successful searches i e walkers that have found the goal node per tick Shortest path distances This plot is plotted when the go dijkstra button is pressed It plots the shortest path distances from each node to the goal node Download free ebooks at bookboon com 58 7 Communication Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication Nodes This m
112. ept and what is a domain Another answer from the Sci Tech Dictionary is A computer system whose usefulness derives primarily from a data base containing human knowledge in a computerized format This definition does not explain what human knowledge is that is it avoids answering the fundamental question at the heart of these questions What is Knowledge A third answer provided by www wikipedia com is as follows Knowledge based systems are artificial intelligent tools working in a narrow domain to provide intelligent decisions with justification Knowledge is acquired and represented using various knowledge representation techniques such as rules frames and scripts The first sentence accurately describes the function of a knowledge base system but not what it is made of The second sentence uses the word knowledge twice without defining what it is A fourth answer can take an agent oriented perspective where the knowledge based system is considered as an agent whose task is to impart its knowledge to the users of the system If information is defined as data i e numbers text images etc that is helpful in answering a question then an agent can be deemed to have knowledge if it knows how to use information to answer a question in a particular context In other words the agent is deemed by us as users of the system to be knowledgeable because it knows how to an
113. er of agents 103 2 see Follow Wanderer number of wanderers 2 ees ees Follow Follower number of followers 51 Follow Avoider number of avoiders 50 Figure 6 7 3 Screenshot of the Interface for the Follow and Avoid model after the setup button has been pressed followed by the go button Download free ebooks at bookboon com 33 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour WHAT IS IT This model attempts to simulate the seeking and fleeing steering behaviours for boids as devised by Craig Reynolds HOW IT WORKS It does this by generating random Wanderer turtles boid that simply wander Followers seeking on detecting Wanderers within their radius turn to face the nearest Wanderer and move forward Avoiders fleeing do the same action but reverse their direction by 180 degrees thus moving away from the nearest Wanderer HOW TO USE IT Setup the initial world Then use the speed bars during operation to adjust the speed of all the turtles THE INTERFACE The buttons in the Interface are defined as follows Setup This clears the environment and creates new Wanderer Follower and Avoider agents Go This starts the simulation Follow Wanderer This will follow one of the Wanderer agents Follow Follower This will follow one of the Follower agents Follow Avoider This will follow one of the Avoider agents The sliders and monitor in the Interface are
114. er to particle agents 1 Run the simulation by then pressing the setup environment button followed by the go simulation button Observe how the particle agents start flowing uphill by somehow managing to tunnel through the surrounding soil nothing like what happens in real life Look at the code that gets executed when this happens See if you can determine what causes the agents to tunnel and try to figure out a way of fixing this in order to make the agents behave the way they might do in real life Download free ebooks at bookboon com 157 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solutions to Selected Exercises Solution to Exercise 6 1 2 The Firebreak model illustrates how the Fire model can be extended to allow the user to add firebreaks extra forest and ignition points as well as rudimentary image processing to simulate a real environment setup go draw forest 2 import forest unburn forest burn all the forestey ignite forest 2 ignite on the left draw firebreak z remove firebreaks percent burned 33 5 a z gt foe I Figure S 6 1 2 Screenshot of the Interface for the Firebreak model after the setup button has been pressed followed by the import forest and burn all the forest buttons with the slider values as shown in the image WHAT IS IT This model is an extension of the Fire model developed by Uri Wilensky tha
115. ering the random rate of turn How does this affect the turtles ability to follow the path Why does it change Change the radius angle and radius length What does this do to the model Why does this change Change the turtle collision parameters How does this change the model What are the optimum settings for accurate collision detection Try altering the turtle speed why does this change the ability to stay in the path Click on the trail button here is the real reason for variable colours EXTENDING THE MODEL Adding a turtles own variable for speed we can make the turtles accelerate and decelerate This will enable a colliding turtle to simply slow down and alter direction to avoid a collision Download free ebooks at bookboon com 28 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour NETLOGO FEATURES The who myself in cone in radius heading sprout and turtle editor features are used in this model RELATED MODELS See the following models Crowd Path Following Flocking With Obstacles Follow and Avoid Obstacle Avoidance 1 Obstacle Avoidance 2 Vision Cone Example 2 Wall Following Example 2 These are basic implementations of various Craig Reynold s steering behaviours for boids Another boid related model is the Biology Flocking model in the Models Library matter what you want out of your future per an_employer with a broad range of operations in a load o
116. ers and switches are defined as follows animate dijkstra This will animate the execution of Dijkstra s algorithm that runs when the go dijkstra button is pressed revisits allowed If set to On this will allow the agents to revisit already visited nodes centre of the universe This sets the goal node that the agents are trying to find It can be set to either Paul Erdos or Kevin Bacon These nodes are randomly created when the network is created g and automatic nomical to create rience expertise and creativity ost performance beyond expectations eed the best employees who can challenge The Power of Knowledge Engineering q A Plug into The Power of Knowledge Engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 56 Artificial Intelligence Exercises Agent Behaviour 7 Communication knowledge sharing This specifies the type of knowledge sharing the agents employ while walking around the network The types of knowledge sharing is as follows None The agents do not exchange any knowledge with other agents As a result the agents end up randomly walking around the network Word of mouth If the agents have found the goal then they share their knowledge with other agents they meet while returning to their start destination Blackboard The agents share their knowledge by writing what they know into communal blackboards that are found at each nod
117. ert Artificial Intelligence Exercises Agent Behaviour 7 Communication 2x2x5x7x3x9991 x100x 20 x 50x 20000 x101 x11x11x 20x 100 x 10x101 2 071793x 10 To run the simulation for all of these settings would not be feasible as the number exceeds the total number of seconds that have elapsed since the beginning of the universe which is estimated at approximately 1 followed by 17 zeros Clearly we have to be more selective of which experimental settings to run in order to investigate the behaviour of the model for the different configurations Thankfully NetLogo provides a tool called BehaviorSpace that makes it easier to explore the space of possible behaviours for this model The tool systematically runs the model multiple times and records the results of each run to a file This process is sometimes called parameter sweeping The tool can be run directly from the Tools menu in NetLogo The way to use it is described in NetLogo s user manual The following excerpt is taken from the manual and explains why such a tool is so useful atter what you want out of your future ber an_employer with a broad range of operations in a load of countries will always be the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our web site www volvogroup com e look rward to g
118. ert Artificial Intelligence Exercises Agent Behaviour 8 Search WHAT IS ITS PURPOSE The purpose of this model is to highlight via basic animation techniques the fundamental differences in the two different distance calculations HOW IT WORKS Two point agents are used to designate the start and end points A walker agent is used to go between the two points This is done either in a direct line to illustrate what the Euclidean distance metric is based upon Or for the Manhattan distance metric it is done through any of the paths that avoid crossing into the Manhattan skyscrapers represented by the blue boxes in the environment In other words the paths can only go down either an avenue roads that run north south or street roads that run west east HOW TO USE IT First press Setup to reset the agents and environment Then press the plot euclidean distance button if you wish to show the path the Euclidean distance metric shown in the Euclidean Distance monitor is based upon Or press the plot manhattan distance button if you wish to plot one of the paths the Manhattan distance metric shown in the Manhattan Distance monitor is based upon Pressing this button again will plot a different path but with the same value shown in the monitor ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on makin
119. es to see what happens with ignition points placed inside and outside of the circles Load the satellite image for Chateaubriant France and start the simulation using different ignition points How well does the simple code for recognizing a tree i e the is tree reporter do at recognizing where the trees in the image are Try different ignition points with different configurations of firebreaks What firebreak drawing behaviour works best at containing the different fires atter what you want out of your future ber an_employer with a broad range of operations in a load of countries will always be the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our web site www volvogroup com e look rward to getting to know you VOLVO AB Volvo publ WWW YONOGTOUD OOM Vowo Teucxs Rexauer Taucks Macx Trucxs Vowo Buses Vowo Coxsteucrion Eoupwent Vowo Pewa Vowo Aero Vowo IT Vowo Financat Sernces Vowo3P Vowo Powenrmam Vowo Parts Vowo Tecumooer VowoLoasncs Business Anca Asu Download free ebooks at bookboon com 162 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Try changing the values of the three sliders to see how this affects the simulation Try pressing the unburn forest button while the fire is s
120. es with less than one year of full time postgraduate work experience London Business School s Masters in Management will expand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School Regent s Park London NW1 4SA Applications are now open for entry in September 2011 United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Download free ebooks at bookboon com 35 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour NETLOGO FEATURES The model uses the command in radius to implement the boids cone of vision RELATED MODELS Another boid related model is the Biology Flocking model Exercise 6 7 4 Obstacle Avoidance 1 NetLogo Model Obstacle http files bookboon com ai Obstacle A voidance 1 html Avoidance 1 NI a gt ticks 671 3D Setup Go al Adjusts the vision of the 1 Bod speed adjusts the step boid sp 0 9 value of each boid per iteration sy es rate of random turn 31 degrees aus ners radius angle
121. ess value The hue is the x dimension and the saturation is the y dimension draw colour circle This draws a colour circle for a specific brightness value The hue values are spread out around the 360 degrees of the circle and the saturation values are spread out along various radii lengths from 0 to 255 from the centre of the circle The model s Interface sliders are defined as follows brightness This controls the brightness of the colour drawn hue increment and saturation increment This controls the increments used when drawing the colours Smaller increments will mean that the drawing has finer detail THINGS TO NOTICE Notice how difficult it is to define the region for a specific colour For example try drawing an imaginary line around everything that is yellow Download free ebooks at bookboon com 122 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge THINGS TO TRY Try using the brightness slider to change the brightness value to see what effect it has on the colour circle and colour rectangle Change the values of the hue increment and saturation increment sliders to change how fine an increment is used to draw the colour circle Increasing the size of the increments will result in symmetrical patterns being drawn Why is this Try doing this while the circle is being drawn You can draw a border to the circle by modifying the following code set pcolo
122. etting to know you VOLVO AB Volvo publ WWW YONOGTOUD OOM Vowo Teucxs Rexauer Taucks Macx Trucxs Vowo Buses Vowo Coxsteucrion Eoupwent Vowo Pewa Vowo Aero Vowo IT Vowo Financat Sernces Vowo3P Vowo Powenrmain Vowo Parts Vowo Tecumooer VowoLoasncs Business Anca Asn Download free ebooks at bookboon com 62 Artificial Intelligence Exercises Agent Behaviour 7 Communication The need for this type of experiment is revealed by the following observations Models often have many settings each of which can take a range of values Together they form what in mathematics is called a parameter space for the model whose dimensions are the number of settings and in which every point is a particular combination of values Running a model with different settings and sometimes even the same ones can lead to drastically different behavior in the system being modeled So how are you to know which particular configuration of values or types of configurations will yield the kind of behavior you are interested in This amounts to the question of where in its huge multi dimension parameter space does your model perform best In this exercise try out the BehaviorSpace tool for the Being Kevin Bacon model in order to gauge which knowledge sharing behaviours perform best at searching the simulated networks First select the BehaviorSpace option form the Tools menu then to create a new experiment setup click
123. ey reflect the way that multiple paths occur leading to different outcomes Is there some way that these procedures can be made more generic and less specific to the problem 8 5 Search as behaviour selection Exercise 8 5 1 Observe situations where search is being employed by a human or an animal including yourself Do you think these searches are best characterized as a choice between alternative sequences of actions or as a choice between alternative sequences of behaviours Or is another characterisation more appropriate Observe the process you employ when you make this decision Exercise 8 5 2 As pointed out in Section 8 5 of the companion Artificial Intelligence Agent Behaviour I book in the Searching Mazes model in order to illustrate the behavioural approach to characterising the search process the agents have been given the ability to execute three different reactive behaviours i moving forward ii turning left then moving forward and iii turning right then moving forward The agents can also execute several different types of forward movement behaviour These are controllable by two choosers in the Interface 1 the move forward behaviour and 2 the move forward behaviour after turning The first defines the forward movement for reactive behaviour i and the second defines the forward movement for the reactive behaviours ii and iii The types of forward movement are as follows a Move forward n steps un
124. f agents to expand the search each step It is only used by the Multi Agent Depth First Search max depth This sets the maximum depth of the search for the depth limited search and the Iterative Deepening Search only The model s Interface monitors are defined as follows Active Searcher Agents This shows the current number of active searcher agents still participating in the search Maximum Active Searcher Agents This shows the maximum value that the Active Searcher Agents monitor has achieved throughout the search Total Searcher Agents This shows the total number of searcher agents that the search has used IDS Depth This is the current depth during the execution of the Iterative Deepening Search THINGS TO NOTICE Notice the effect of turning the flag allow revisited states On and Off Why is turning it Of f so effective at dramatically reducing the search for the different search behaviours Download free ebooks at bookboon com 88 Please click the advert Artificial Intelligence Exercises Agent Behaviour 8 Search Notice that some of the searches if repeated with the same settings end up in different places Why is this Notice that many of the behaviours result in the searching turtle agents getting stuck When does this happen and why THINGS TO TRY Try slowing down the animation by changing the speed slider at the top of the Information window You will be able to see how the search pr
125. f countries will always be the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our el Site www volvogroup com e look forward to getting to know you VOLVO AB Volvo pub WWW YONOGTOUD COM Vowo Teucxs Rexauer Trucns Macx Trucxs Vowo Buses Vowo Coxsteucrion Eouipwext Vowo Pena Vowo Aero Vowo IT Vowo Financat Sernces Vowo3P Vowo Powenrmam Vowo Parts Vowo Tecumooer VowoLoasncs Business Anca Asu Download free ebooks at bookboon com 29 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Exercise 6 7 2 Flocking with Obstacles NetLogo Model Draw and Draw T erase collision Eraser patches Load an image overlay then chose the correct collision patch Use the draw and erase editor to alter the images or create your own Figure 6 7 2 Screenshot of the Interface for the m with Obstacles model after the setup button has been pressed followed by the go button WHAT IS IT This model is an attempt to recreate the flocking behaviour of Craig Reynold s boids It was created by Uri Wilensky and modified by Thomas Christy and William John Teahan Most of the Information in the model is similar to that provided by Uri Wilensky with some modifications as included below
126. f the path has been ended for example by the end path command In this case if you do not wish to erase everything by pressing the setup command then the best strategy is to draw over the top of the path with another path of a different colour to the desired background colour There are two other types of paths these are paths drawn using patch agents using the draw patches path command and using turtle agents using the draw turtles path command Unfortunately the quality of these types of paths can be very patchy excusing the pun Only the turtle agents type of path can be selected then moved or erased The patch type of path cannot be selected unlike the path drawn using the turtle s pen described above The model uses drawer turtle agents for drawing the paths tracer turtle agents for tracing them text turtle agents for drawing the text and square drawers for drawing a path using square turtles HOW TO USE IT Press the setup button first To place an object in the map first select the desired object type and object colour press the draw object button then place the object at the desired Download free ebooks at bookboon com 127 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge location by moving the mouse and then clicking You can erase the object by pressing the erase object button then clicking on it with the mouse To move it press the select object button then when it is highlighted
127. g the world a better place for people business and society Join us In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To apply and for all current job openings please visit our web page www ericsson com careers Download free ebooks at bookboon com 111 Artificial Intelligence Exercises Agent Behaviour 8 Search THE INTERFACE The model s Interface buttons are defined as follows Setup This resets the model plot euclidean distance This draws the straight line path for which the Euclidean distance calculation is based upon plot manhattan distance This draws one of the paths for which the Manhattan distance calculation is based upon The model s Interface monitors are defined as follows Euclidean Distance This is the Euclidean distance between the two points Manhattan Distance This is the Manhattan distance between the two points THINGS TO NOTICE Notice that Euclidean distance is always less than Manhattan distance Why Both Euclidean distance and Manhattan distance are admissible heuristics if used for an informed search algorithm such as greedy search or A search Why THINGS TO TRY Keep plotting the different Manhattan paths by repeatedly pressing the plot manhattan distance button A visually appealing pattern will emerge Some of the
128. g walkers Note also how the next locations list parameter for the expand walkers command determines the knowledge sharing strategy being used that defines the different searches RELATED MODELS See the Searching for Kevin Bacon model 7 6 Using communicating agents for searching networks Exercise 7 6 1 In the Being Kevin Bacon model there are many different slider chooser and switch values to try out when running the search simulation This is done by first pressing the setup network button followed by the go walkers button For example there are five different knowledge sharing methods None Word of mouth Blackboard Combined 1 and Combined 2 and there are also seven different types of networks P2P no supernodes P2P has super nodes P2P random single link P2P 33 ce incremental P2P incremental 1 Star central hub and Hierarchical Indeed if we wished to explore all the different configurations for the search simulation then moving through all the switch chooser slider values from right to left and from top to bottom i e those for all the green Interface elements but ignoring the two possible settings for the animate dijkstra switch as that switches between Dijsktra s algorithm and the search simulation then the total number of different settings would be as follows Download free ebooks at bookboon com 61 Please click the adv
129. global behaviour of the colony or flock when you are watching a single turtle zoomed most of the way in or when zoomed all of the way in i e when you are in ride mode Slow down the speed of the simulation so that you verify the local behaviour of the turtle you are watching from a first person perspective Conversely how easy is to discern the local behaviour of each turtle when you are no longer watching one but are the observer looking at the world from above 6 5 Stigmergy and Swarm Intelligence Exercise 6 5 1 Which of the following models in NetLogo s Models Library are examples of swarm intelligence where the agents make use of stigmergic local knowledge Download free ebooks at bookboon com 18 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Flocking In NetLogo s Models Library Biology gt Flocking http ccl northwestern edu netlogo models Flocking Moths In NetLogo s Models Library Biology gt Moths http ccl northwestern edu netlogo models Moths Termites In NetLogo s Models Library Biology gt Termites http ccLnorthwestern edu netlogo models Termites Virus In NetLogo s Models Library Biology gt Virus http ccl northwestern edu netlogo models Virus 6 6 Implementing behaviour of Turtle Agents in NetLogo Exercise 6 6 1 Heatbugs NetLogo Model Try out the Heatbugs model in NetLogo Heatbugs In NetLogo s Models Librar
130. gt lt colour gt This sets the count for the specified colour in the list of counts maintained for the agent whose name is lt agent name gt These counts are used to determine the probability for a specific colour using the following formula P colour C colour C total where C colour is the count for the symbol and Cora is the sum of all counts for all colours total The model s Interface monitors are defined as follows adaptive lt colours This is the count of the number of cars observed in the animation of the respective colour that is adaptively updated as the animation proceeds lt agent name gt Entropy This is the entropy of the probability distribution maintained by the agent whose name is lt agent name gt lt agent name gt Code Length This is the cost of optimally encoding the sequence of observed cars given the agent s probability distribution The model s plot is defined as follows The distributions are converted to a number that uniquely represents the values of the counts in the distribution For example the number 1234 can be thought of as representing four separate counts 1 2 3 and 4 that combine to produce a unique number In this case the separate counts do not range from 0 to 9 they can range from 0 to 1000 The number that uniquely represents the distribution is then plotted versus ticks to show how the distributions evolve over the simulation Download free ebooks
131. hat arise with representing the real world Once finished check how many errors there are and how much detail has been missed off the map A month or so latter check again and see how much has changed Try loading the Central Park map This is done by using the NetLogo menu option for importing a world Select the File menu from the NetLogo menu bar then select Import World option Download free ebooks at bookboon com 131 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge EXTENDING THE MODEL This model has a long way to go before it could be used to draw more high quality maps such as that used for orienteering One of the main drawbacks is the poor quality of the path drawing Can this be improved in some manner NETLOGO FEATURES The Import World and Export World menu options are used to save the map to disk and re load it Exercise 9 7 2 Adapt the code from the Map Drawing model so that you can annotate any imported map or image with the following e circles e arrows e text e marks an x for marking a spot Try this A Ki options The sequence 2 4 6 8 10 12 14 16 Is The the SOG ENCE of even Whole Numbecs l iS the NUNbEr 00th place in this Sequence Challenging Not challenging Try more gt www alloptions nl life Download free ebooks at bookboon com 132 Artificial Intelligence Exercises Agent Behaviou
132. he boids cone of vision boid speed This specifies how far the boids should move each tick boid bounce This specifies how far the boids should bounce back when they hit the tank walls or the earth Download free ebooks at bookboon com 154 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence 3 For when the solution variable is set to particle agents 1 max particles This specifies the maximum number of particles and can be used to control how many particles come out of the tap initial velocity x and initial velocity y This specifies the initial velocity of the particle in the x and y directions initial range x This provides some random variability in the velocity in the x direction restitution constant This can be used to control how much bounce with floor damping occurs if the particle touches the tank floor gravity constant This is used to scale the force of the gravity according to the particle s mass thereby simulating the effect of air resistance viscosity constant This can be used to simulate the effect of viscosity rate This controls the rate of particle emission step size The new location and speed of each particle is calculated by multiplying the forces by this number The ticks also advance by this number wind velocity x and wind velocity y This specifies the velocity of the wind in the x and y directions THINGS TO NOTICE When the chooser so
133. hen load the following data file for Exercises 8 3 1 and 8 6 2 Set the search behaviour chooser to the relevant search algorithm Also set the value of the start node number slider to 1 and make sure the output trace switch is set to On For the informed searches set the heuristic chooser to Manhattan Distance ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on making the world a better place for people business and society Join us In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To apply and for all current job openings please visit our web page www ericsson com careers Download free ebooks at bookboon com 177 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Descriptions of this exercise for each search algorithm can be generated using the Generate Problem Description button Then to generate a trace of the search as it proceeds first press the setup searcher button and then press the go searchers button The output produced for the Breadth First Search on Network 1 is as follows Figure 8 3 1 1 Screenshot of the Environment for the Searching for Kevin Bacon 2 model showing how the Breadth First Search se
134. here inside the rectangular box then drag it and everything inside it using the mouse All the objects inside the box will move to the point where the mouse is dragged to The Interface s choosers sliders and input box are defined as follows map or image to use This specifies which image gets loaded into the environment If set to User specified then the user will be prompted for an image filename This will not work when the model is invoked in an applet arrow colour This sets the colour of the arrow that is drawn when the draw arrow button is next pressed The colour is specified in a range from 0 to 140 using NetLogo s colour numbering scheme Sample colour numbers are shown in the small text note at the left bottom of the Interface arrow start size This sets the size of the arrow tail drawn as a filled in circle arrow end size This sets the size of the arrow head drawn using the default arrowhead shape arrow thickness This sets the thickness of the arrow s shaft Download free ebooks at bookboon com 190 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises circle colour This sets the colour of the circle that is drawn when the draw circle button is next pressed The colour is specified in a range from 0 to 140 using NetLogo s colour numbering scheme Sample colour numbers are shown in the small text note at the left bottom of the Interface cir
135. how various ial features of ea models and to visualise them using NetLogo link and turtle agents h S angua Map Drawing O71 Users can create their own maps using this model Mazes 2 6 1 1 This model extends the Mazes model It gets a simple reactive turtle agent to move around several mazes s bc 1 AZES Missionaries and Cn e 2 dy This model applies standard search algorithms to the classic search problem called Missionaries and Cannibals bookbe 28 lt a als Moths 6 5 1 Sa to 6 5 D This model demonstrates moths flying around a light in circles each following a set of simple rules In NetLogo s Models Library Biology gt Moths Obstacle Avoidance 1 6 7 4 This model is an attempt to recreate boids Ce ee Reynold s work that ye basic obstacle avoidance steering behaviour http files bookboon co Obstacle A voidance y Obstacle Avoidance 2 6 7 5 This model is an attempt to recreate boids eee Craig Ronee s aan that o beste obstacle avoidance steering behaviour http files bookboon com ai Searching for Kevin Bacon 8 2 3 This model eae standard search ae to pe Pen of E for a specific goal node in a network files bookbe m ai Searching for n Bacon ht Searching ian Kenn Bacon 2 2 82 DAL BS 6 2 3 7 il solutions 08l 8 O25 Bodo ll This model is an extension to the Searching for Kevin Bacon model It provides an Output box that allows the user to trace how the search procee
136. il it reaches the goal state or other termination criteria are satisfied i e it reaches the right side of the graph without finding a solution Pressing the go animation button will start the animation and show the path taken at the same time in the graph This is shown using the red colour Download free ebooks at bookboon com 93 Artificial Intelligence Exercises Agent Behaviour 8 Search THE INTERFACE The model s Interface buttons are defined as follows setup This will clear the environment and all variables reset the animation and redraw the graph go once This will make the search proceed one step at a time go forever This will make the search proceed continuously until it has reached the goal state or it is deemed to be unsuccessful go animation This will start the animation shown at the middle bottom of the Interface The model s Interface choosers sliders and switch are defined as follows search behaviour This specifies what search strategy the agents should employ when performing the search A full description of the different search strategies can be found in Chapter 8 of the book Artificial Intelligence Agent Behaviour P heuristic This specifies the heuristic to be used for the Informed searches Greedy Best First Search and A Search max depth This sets the maximum depth of the search max agents to expand This sets the maximum number of agents to expand th
137. initialise the search and create a random network whose type is specified by the network type chooser press the setup network button in the Interface To create your own network set the network type chooser to User specified If the specify coordinates slider value is set to Off then the user can then use the mouse to click on the points in the environment where they want the network nodes located The first point clicked will be designated as the goal node i e the Kevin Bacon node If the slider is set to On then the user will be prompted for the x and y co ordinates of the nodes For each of the nodes in the network the user will then be asked how many links there are from the node and which nodes link to it Yow re full of energy and ideas And that s just what we are looking for UBS 2010 All rights reserved Looking for a career where your ideas could really make a difference UBS s Graduate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates www ubs com graduates a UB S Download free ebooks at bookboon com 104 Artificial Intelligence Exercises Agent Behaviour 8 Search To save the network to a file press the Save Network to a File button To reload it a
138. ions pose problems that require a re evaluation of the suitability of your taxonomical classification Download free ebooks at bookboon com 117 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Exercise 9 1 2 What must a knowledge based system have that is different to a database system in order for it to become a repository of knowledge rather than just a repository of data Your answer must include a discussion of the following claim To create a knowledge based system all you need to do is add some rules to a database system to infer new data 9 2 Knowledge as justified true belief Exercise 9 2 1 Discuss the following statement Knowledge cannot exist by itself independent of us It only exists in our heads Exercise 9 2 2 Discuss the standard definition of knowledge that states that knowledge is justified true belief What types of knowledge are well suited to being characterized in this manner What types of knowledge pose problems for this definition 9 3 Different types of knowledge Exercise 9 3 1 Label each of the following as being declarative knowledge procedural knowledge task knowledge behavioural knowledge episodic knowledge explanatory knowledge or inferred knowledge e the knowledge that Margaret Thatcher is still the current prime minister of Great Britain e the knowledge that the Titanic sank on April 15 1912 e the knowledge of how to cook
139. ired to explain the meaning of life the universe and everything e the knowledge that the sun will explode at the end of this century in the year 2099 e the knowledge that global warming is causing irreversible damage to the Earth s environment e the knowledge that you think are self aware and are conscious Which of these do not fit well in any of the categories What problems arise when trying to represent the knowledge encompassed by each of these in a computer system 9 4 Some approaches to Knowledge Representation and Al Exercise 9 4 1 Compare and contrast the following approaches to knowledge representation e symbolic e connectionism non symbolic e conceptual spaces Exercise 9 4 2 Discuss for each of the three approaches listed in Exercise 9 4 1 above how you might design the knowledge to describe the following e the sound of bagpipes e g to a deaf person e the taste of milk e what sandpaper feels like Download free ebooks at bookboon com 119 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge e what coffee smells like e what your face looks like e g to a blind person For each approach indicate the shortcomings in the solution you chose Exercise 9 4 3 Colour Cylinder NetLogo Model Try out the Colour Cylinder model in NetLogo In Paris or Online International programs taught by professors and professionals from all over the w
140. ironment that is moving around following the walls NETLOGO FEATURES The model uses the layout spring command for modifying the layout of the network of states turtle agents and paths link agents ience expertise and creativity erformance beyond expectations Plug into The Power of Knowledge engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 23 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour RELATED MODELS See the Wall Following Example 2 model the Central Park Events model and the Knowledge Representation model For the Wall Following Example 2 model see Exercise 6 7 6 below 6 7 Boids Exercises 6 7 1 to 6 7 6 Try out the following models that implement various steering behaviours for boids 6 7 1 Crowd Path Following 6 7 2 Flocking with Obstacles 6 7 3 Follow and Avoid 6 7 4 Obstacle Avoidance 1 6 7 5 Obstacle Avoidance 2 6 7 6 Wall Following Example 2 Note that to enhance the boids simulation and to be fully compatible with Craig Reynold s boids implementations for these models this requires the addition of gradual acceleration and deceleration A faithful implementation should also include some form of steering force implemented using point mass approximation where each boid has a mass and works in relation to forces However these NetLogo implementations show how an approximation to Reynolds approach can be achieved relatively easy a
141. is to choose all in random order then the three modified behaviours see above will be executed together but in no particular order Otherwise just one of them will be randomly chosen to be executed number of turtles This is the number of turtle agents that are created at setup THINGS TO NOTICE Changing the behaviour variable from Standard to Modified in the Interface does not seem to change what the turtles seem to be doing However when the behaviour is set to Modified see if you can notice slight variations in behaviour to the Standard behaviour especially when the turtle agent ends up in a cul de sac or there is a narrow alley way between two buildings In the latter case the Modified behaviour results in the turtles having the ability to explore further than their Standard counterparts Download free ebooks at bookboon com 45 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour When the behaviour variable is set to Modified note that the speed of the turtles is much faster when the choose sub behaviours variable is set to choose all in random order rather than choose one at random Why Exercise 6 7 7 Look at the code in the NetLogo models described in Exercises 6 7 1 to 6 7 6 How is each of the models performing obstacle avoidance i e Locate and explain the piece of code in the models that performs obstacle avoidance Or are there any models not performing any obstacle avoidance at all Exer
142. ises Il Exerc John Teahan S Q gt ies O aa i D O lt O a D _ i O mu PSE n William Teahan Artificial Intelligence Exercises Agent Behaviour Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour 2010 William Teahan amp Ventus Publishing ApS ISBN 978 87 7681 592 9 Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Contents Contents Preface 8 6 Behaviour 13 6 1 What is behaviour 13 6 2 Reactive versus Cognitive Agents 15 6 3 Emergence Self organisation Adaptivity and Evolution 16 6 4 The Frame of Reference Problem 17 6 5 Stigmergy and Swarm Intelligence 18 6 6 Implementing behaviour of Turtle Agents in NetLogo 19 6 7 Boids 24 Te Communication 47 7 1 Communication Information and Language 47 de The diversity of human language 47 Lo Communication via communities of agents 48 7 4 Communicating Behaviour 49 T The Small World Phenomenon and Dijkstra s algorithm 51 7 6 Using communicating agents for searching networks 61 Ted Entropy and Information 66 7 8 Calculating Entropy in NetLogo 66 7 9 Language Modelling 70 7 10 Entropy of a Language 76 7 11 Communicating Meaning 81 Fast track your career Masters in Management Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experi
143. it symbol 5 probability 1 12 negative log probabilit y y y y y y Entropy of distribution 17 603 Figure S 7 7 1 Screenshot of the Interface for the Entropy Calculator model after the setup button has been pressed followed by the calculate entropy button with the slider values as shown in the image 6007 a g swasis yay 20u It s only an opportunity if you act on it IKEA SE STUDENT ges aa Download free ebooks at bookboon com 174 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises WHAT IS IT This model allows the user to calculate the entropy for a specific probability distribution The analogy is that there is an imaginary bag that a person pulls coloured balls from The range of colours that the ball can have is its alphabet Each colour is a symbol Each symbol has a frequency count which can be used to determine its probability by dividing the count by the sum of the counts from all the symbols The probabilities therefore sum to 1 and form the probability distribution for the set of balls Entropy provides a measure of how much variation there is in the probability distribution It will be minimum when the symbols are all equiprobable If one of the symbols have a higher count than the others then the entropy will be correspondingly higher than the minimum when all the counts are equal WHAT IS ITS PURPOS
144. it queue data structure separate from what is happening in the search is not needed although technically the information is still being stored in a queue behind the scenes but the implementation of this is hidden behind the top level commands that NetLogo provides SimCorp is a leading provider of software solutions for the financial industry We work together to reach a common goal to help our clients succeed by providing a strong scalable IT platform that enables growth while mitigating risk and reducing cost At SimCorp we value commitment and enable you to make the most of your ambitions and potential Are you among the best qualified in finance economics IT or mathematics m www simcorp com S l m Co rp MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 92 Artificial Intelligence Exercises Agent Behaviour 8 Search The hope is that this queueless agent oriented approach provides a more intuitive solution that makes it easier to grasp how the search strategies work It uses a situated embodied first person design perspective therefore arguably making it easier to understand the essential differences between the search strategies The breed of searcher turtle agents is used for the searchers that move throughout the maze trying to get to the goal These searcher agents maintain information about the current state of the search such as the time taken and the path and estimat
145. itations of the approach that the reactive agent uses compared to the cognitive agent Download free ebooks at bookboon com 15 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour 6 3 Emergence Self organisation Adaptivity and Evolution Exercise 6 3 1 Try out the following models in NetLogo Explain for each of these models whether you witness any of the following e Reactive behaviour in the turtles e Cognitive behaviour in the turtles e Emergence e Self organisation In order to answer this question fill in the rest of the table provided below Hypothetical answers are provided for the Hypothetical model which does not exist in the table below Exercise 6 3 2 Try creating a model in NetLogo to illustrate continental drift between Australia and New Zealand Hint Use the import pcolors command to import an image containing the present shapes of Australia and New Zealand Add a monitor into the model to show an estimate of the years passed as the two countries move slowly back together Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Exercise 6 3 3 Is there some way of measuring self organisation when it appears in a model For example for the Termites and State Machine Example models is there some way of measuring how successful the turtle agents have been at moving the woodchips into
146. l NETLOGO FEATURES Note the use of the hatch and die commands to clone child searchers and terminate parent searchers thus avoiding the need for a separate queue data structure to maintain the information during the search Note also the variation in the ask command that determines the strategy being used that defines the different searches RELATED MODELS For other search problems see the Searching For Kevin Bacon model and the Searching Mazes model For an insight into the Manhattan distance and Euclidean distance metrics used as heuristics for the Informed searches see the Manhattan Distance model Exercise 8 2 3 Searching for Kevin Bacon NetLogo Model Try out the Searching for Kevin Bacon model in NetLogo Download free ebooks at bookboon com 96 Artificial Intelligence Exercises Agent Behaviour 8 Search ay setup network setup searcher Ady ticks 7 ao change layout o searchers once iinan MOR ay ige lay g b80 b7 bi3 b27 b57 b83 reset searchers go searchers m b39 b17 b25 b4 network type layout type b76 h34 bs bJ9 P2P incremental v spring v big oon I b99 E Ee AX J 20 a ns X Eee 42 4 z big b8 search behaviour start node number b6 b2 b33 b86 Hill Hill Climbing Search Y Search ia i b94 b91 bla pzz heuristic b6S fary Bs ke Euclidean distance y b52Cb93 ba ber b23 b102 max agents to expand 2 b30 bs1 bH2 aon allow revisited nodes bay b107 574 w b55 b82
147. ld specifies which reporters in the model are executed in order to collect data about the run These are recorded in the output file that is produced The Measure runs at every step ensures that these reporters are executed every step The setup commands field specifies which commands in the model are executed at the start of each run and the Go commands field specifies the commands that are executed each step There are also fields for specifying the stop condition for the run the final commands to be executed at the end of the run and a fixed time limit if set to 0 no time limit is specified Once you have filled in the Experiment dialog box press OK at the bottom then press Run in the BehaviorSpace dialog box A Run options dialog box will then appear If you wish to have the results of the experiment saved to a file with a format suitable for loading into a spreadsheet application then tick the Spreadsheet output option followed by the OK button An Exporting as spreadsheet dialog box will appear where you can specify the name of the results output file Pressing the Save button will then start the experiment running Your Interface should end up similar to that shown in Figure 7 6 1 Once the experiment is finished you can run it again by pressing the Run button in the Experiment dialog box Download free ebooks at bookboon com 64 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Com
148. le 2 model for further explanation The events are shown using an event map representation where events are linked to other events on separate streams in an ordered sequence The idea is that an agent processes events simultaneously on separate streams The approach is similar to the approach adopted for Event Stream Processing ESP WHAT IS ITS PURPOSE The purpose of this model is to show how to visualise a series of events using an event map Download free ebooks at bookboon com 21 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour HOW IT WORKS The model uses turtle agents to represent the states in the event map and uses links agents to represent the paths between states States own three variables depth The depth in the event map tree stream The stream name where a stream consists of a sequence of sensory or motor events event The event either sensory or motor A spring layout is used to visualise the event map HOW TO USE IT Press the setup button first This will often produce a cluttered layout To unfold the clutter press the change layout button and then dynamically change the values in the sliders that control the layout One effective technique is to reduce the value of the spring length slider to 0 then slowly increase it back up again until the desired length is achieved THE INTERFACE The model s Interface buttons are defined as follows setup This will clea
149. less hit wall where the agent moves forward n steps and n is a number defined by the Interface variable move forward step amount b Move forward until hit wall where the agent will keep on moving forward until it hits a wall c Move forward until any open space where the agent will move forward until there is a wall in the way or there is open space on both sides of it and d Move forward until open space after side wall where the agent will move forward until there is a wall in the way or there is open space after a side wall Download free ebooks at bookboon com 108 Artificial Intelligence Exercises Agent Behaviour 8 Search Despite only being able to execute three basic reactive behaviours the maze searching agents can exhibit a surprising variety of movements as a result Figure 8 10 captures a few of these in screenshots taken from the Searching Mazes model using breadth first search on the empty maze Referring to the numbers and letters in the previous paragraph to describe the four configurations in the top left image the agents apply behaviours 1 a and 2 a both using a step of 10 in the top right image the behaviours are 1 a using a step of 1 and 2 b in the bottom left image the behaviours are 1 a using a step of 10 and 2 c and in the bottom right image the behaviours are 1 c and 2 a using a step of 10 For example the grid effect in the top left image is caus
150. lied by the number of ticks is the number that appears in the Years ago monitor NZ stamp ticks modulus This determines how many times the shape of New Zealand is stamped during the simulation When the remainder of the number of ticks divided by this number is 0 then all New Zealand agents i e the agents coloured black will be stamped thereby leaving an impression in the environment of their current position at the time of the stamping This can be used to give an impression of where New Zealand has been at earlier stages of the animation NZ heading This is the heading of New Zealand as it moves towards Australia Changing this to another value will result in New Zealand heading elsewhere NZ forward increment This controls how far New Zealand moves forward each tick on the heading specified by the NZ heading slider Fast track your career Masters in Management Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experience London Business School s Masters in Management will expand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School
151. lt symbol gt lt number gt lt colour gt This sets the count and colour for the specified symbol If a number of the symbol is gt the alphabet size then these symbols are ignored in the calculations These counts are used to determine the probability for a specific symbol using the following formula P symbol C symbol y C total where C symbol is the count for the symbol and C is the sum of all counts for all symbols The model s Output is used to show how the entropy value is calculated THINGS TO TRY Try changing the distribution counts in the sliders to see what affect this has on the entropy calculations Try to find out the conditions when the entropy is minimised Then try to determine if there is a maximum entropy value To do this edit one of the sliders to increase the maximum frequency count for a particular symbol Increase the maximum count from 100 to 1000 What does this do to the entropy as a result Then increase it to 1000000 What happens Download free ebooks at bookboon com 176 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 7 7 2 Minimum entropy occurs when all the symbols are equiprobable Solutions to Exercise 8 3 1 8 6 2 and 8 7 1 Solution to these exercises can be obtained by running the Searching For Kevin Bacon 2 model Press the Load Network from a File button in the Interface t
152. lution in the Interface is set to turtle agents 1 the agents behave a bit like water but do not spread out at the bottom Why How can their behaviour be changed to fix the problem When the chooser solution in the Interface is set to turtle agents 1 the agents behave a bit like snow Why How can their behaviour be changed so that they behave more like water When the chooser solution in the Interface is set to patch agents 1 the agents do not flow correctly on the right Why How can their behaviour be changed so that the right side of the reservoir fills up correctly and then flows via the connecting pipe to the next reservoir and so on When the chooser solution in the Interface is set to patch agents 2 patch agents 3 or patch agents 4 the simulations are complete failures Can any of these be salvaged in order to find a possible solution When the chooser solution in the Interface is set to boid agents 1 the agents behave a bit like spray painting in some configurations Why Also like the turtle agents 1 solution they do not spread out at the bottom Why How can their behaviour be changed to fix the problem How can their behaviour be changed so that they behave more like water When the chooser solution in the Interface is set to particle agents 1 the agents tunnel through the surrounding earth Why How can their behaviour be changed so that they behave more like water Dow
153. lutions 1 and 2 Used by boid agents solution 1 i Figure 10 8 1 Screenshot of the Interface for the Water Flowing Uphill model after the setup environment button has been pressed followed by the go simulation button The solution selected by the solution chooser is boid agents 1 WHAT IS IT This model tries to visually simulate one possible solution to the problem of trying to get water to flow uphill The solution is to use tanks or reservoirs at increasing heights and have these fill up one after the other A stop valve shown in blue at the end of the connecting pipes between Download free ebooks at bookboon com 152 Please click the advert Artificial Intelligence Exercises Agent Behaviour 10 Intelligence reservoirs in the model is used to prevent back filling of the reservoirs once the water moves through the valve it cannot flow backwards Various methods are tried to simulate how the water flows All of these methods are failures some of them producing strange effects that are impossible in the real world The reader is encouraged to add his or her own methods which better approximate how water flows in real life WHAT IS ITS PURPOSE The purpose of the model is to draw an analogy between a seemingly impossible task trying to get water to flow uphill and what some people claim to also be impossible creating Artificial Intelligence For both of these tasks w
154. ments a turtle with a cone of vision INTERFACE The model s Interface buttons are defined as follows Setup This sets up the environment with a grid of obstacles and an outside border drawn using blue patches One turtle agent the wanderer boid is created and placed at a random location Go The boid starts wandering around the environment avoiding obstacles Download free ebooks at bookboon com 40 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Draw Obstacle The user can draw further obstacles in the environment These are coloured brown Follow Wanderer This allows the perspective of the visualisation to be altered so that it is centred on the wanderer Plot This instructs the wanderer turtle agent to put its pen down when wandering Hence this will draw the path it has taken while wandering around The model s Interface sliders are defined as follows boid speed This controls the speed of the boid i e how much it moves forward each tick vrate of random turn This controls how much the wandering boid turns each time tick The boid has a tendency to head in a right turning direction as the rate of random turn to the right as specified by the slider is twice that of the rate of random turn to the left goo n a swasis yar 20u aus It s only an opportunity if you act on it IKEA SE STUDENT ev Download free ebooks at
155. models in order to overcome the zero frequency problem This is the problem that occurs when an upcoming character has not been seen before in the context and therefore has zero frequency Assigning a zero probability estimate will result in an infinite code length since log0 00 which means that it is impossible to encode the occurrence using this estimate To overcome this the estimates are calculated by smoothing with lower order models using the escape mechanism The escape probability is estimated using Method C devised by Alistair Moffat hence why it is called PPMC The smoothing method is often erroneously called Witten Bell smoothing although it was invented by Alistair Moffat The PPMC language model implemented by this NetLogo model does not implement full exclusions or update exclusions WHAT IS ITS PURPOSE The purpose of this NetLogo model is show how to implement a language model based on the PPMC compression scheme and then use it to predict text one character at a time Download free ebooks at bookboon com 78 Artificial Intelligence Exercises Agent Behaviour 7 Communication HOW IT WORKS The data being modelled is conceptualised as a stream of events i e analogously to Event Stream Processing or ESP where events occur with a specific order for a specific stream but at the same time can occur simultaneously on separate streams A training text sequence is used to train the language model This langu
156. ms Enumerating the full knowledge involved and scaling up to a non trivial real world problem presents many difficulties and is still an active area of research in Artificial Intelligence However the merit of these examples is their simplicity and their usefulness in illustrating how the models differ What would need to be done to scale up the knowledge for these problems to their real world equivalents for example identifying the name of an animal while walking around a zoo or the Download free ebooks at bookboon com 142 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge name of a New Zealand bird while out tramping the Milford Track or identifying the different types of yachts seen in a harbour Exercise 9 8 5 Explain the difference between forward reasoning and backward reasoning when rules are chosen as the KR type for the Knowledge Representation model Hint One way of answering this question is by looking to see how they are implemented in the NetLogo code for the model 9 9 Representing knowledge using rules and logic Exercise 9 9 1 Consider the following sentences Peter is an old man Mary is the sister of Peter Some cats have large ears If X is the sister of Y and Y is a boy then Y is the brother of X Simba is a white elephant All elephants are mammals Elephants have a trunk and large ears Ruby is a black and white cat Ruby is the mother of Pearl and Pearl is the d
157. munication BehaviorSpace Interface Information Procedures VY view puate I z Experiments as si a ane Settings t e un licks t experiment 20 runs E _ a ticks 1312 3D New Edit Duplicate Delete Run f 30 Running Experiment experiment ss 1 normal speed M report success o Behavor amp 0 Time 1400 Total elapsed time 0 00 41 links per node 5 layout type circle animate dijkstra true nodes to delete 0 centre of the universe Kevin Bacon new walkers per node 1 simulation ticks 5000 naw walkarc nor tick M Update view IVi Update plots and monitors Run 1 of 20 step 1292 Abort Figure 7 6 1 Screenshot of the Interface for the Being Kevin Bacon model with the BehaviorSpace tool in action using the settings above but with the no of nodes set at 100 rather than 1000 Try this HF 1 options The Sequence D A amp 1 i 0 12 fe 16 5 The iS the NUNbEr the SOG vince of even Whole Numbecs looth place in this Sequence Challenging Not challenging Try more gt www alloptions nl life Download free ebooks at bookboon com 65 Artificial Intelligence Exercises Agent Behaviour 7 Communication Have a look at the results of your experiment by loading the output file that was generated into a spreadsheet Analyse the results to find out which of the knowledge sharing methods
158. nd in many cases the resultant behaviour of the boids is as desired Exercise 6 7 1 Crowd Path Following NetLogo Model Download free ebooks at bookboon com 24 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour E 8 Initialisation 1 m Turtle Movement i Turtle Collision m Patch Collision i im I i Figure 6 7 1 Screenshot of the Interface for the Crowd Path Following model after the setup button has been pressed followed by the go button when the behaviour has been set to Basic crowd following WHAT IS IT This model is an attempt to recreate boids see Craig Reynold s work that use the crowd path following steering behaviour HOW IT WORKS Each turtle boid has two levels of vision a forward facing cone and a boundary radius The Cone looks out for coloured patches i e not black in front and if a patch is detected it runs a simple routine to navigate the path The radius detection simply looks out for other turtles at a given distance threshold When a turtle bumps into another turtle it does not move forward but instead turns right this causes a slight pause in its forward motion which prevents it from going through the colliding turtle As an addition to highlight the collision effect each turtle is given a variable speed based on its who number This makes the first turtle the slowest and the last the f
159. new searcher 19 at node nO Path cost estimated cost combined cost of searcher 19 60 0 60 Searcher agent 19 has found the goal node n0 The output produced for the Hill Climbing Search on Network 2 is as follows Figure 8 3 1 5 Screenshot of the Environment for the Searching for Kevin Bacon 2 model showing how the Hill Climbing Search using the heuristic Manhattan Distance searches the network shown loaded from the Network 2 data file The searcher agent 21 is starting at node 16 Expanding the search for searcher 21 by hatching new searcher 22 at node n11 Height of searcher 22 35 Expanding the search for searcher 21 by hatching new searcher 23 at node n17 Height of searcher 23 30 Expanding the search for searcher 23 by hatching new searcher 24 at node n12 Height of searcher 24 25 Expanding the search for searcher 23 by hatching new searcher 25 at node n16 Height of searcher 25 40 Expanding the search for searcher 23 by hatching new searcher 26 at node n18 Height of searcher 26 20 Expanding the search for searcher 26 by hatching new searcher 27 at node n13 Download free ebooks at bookboon com 182 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Height of searcher 27 15 Expanding the search for searcher 26 by hatching new searcher 28 at node n17 Height of searcher 28 30 Expanding the search for searcher 26 by hatching new sear
160. nges the layout to move the agents apart if the Layout type slider is set to spring go walkers If a language model is visualised in the environment then this will run an animation that shows how the testing sequence specified within the test text input box is processed using the language model kill all walkers This kills all current walkers load model from text This will load a language model using the training text specified by the which text chooser The type of events characters or words is specified by the text event chooser The NetLogo model s Interface sliders and choosers are defined as follows layout type This specifies the type of layout to use for the visualisation either spring or radial which text This specifies the training text to use to train the language models text event This specifies the type of event in the training text being processed either characters or words Download free ebooks at bookboon com 73 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication max tree depth This specifies the maximum tree depth of the language model the order of the language model is equal to the max tree depth 1 max text size This specifies the maximum number of characters to be processed from the training text A relatively low number is recommended otherwise there will be too much to visualise window size This sets
161. nity if you act on it IKEA SE STUDENT ev Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour Exercises for Artificial Intelligence Agent Behaviour k A ie yy i zi Tra af a The Lake ar L Bethesda e m Fields amp ics d QSummer E 7 th stage St wg Skates i eo Ron Tre 7 aie eee aq L Meadow eB Bias Friedsam le ans il Carousel e m ei li Heckscher eiman Ge Playground in E q 1 The Pond m Columbus ae a x ye Central Park South 59th St ies a Selection of screenshots taken from NetLogo models described in this book Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour Preface Preface The list of exercises chapter headings and section and NetLogo models in this book closely follow what is in the companion Artificial Intelligence Agent Behaviour I book The best way to learn about what is written in the companion book is to try out each of the NetLogo models that are described in the book and in the exercises below An index of the models used in these books can be found using the following URL A table listing all the models described in this book and the companion book is also provided below Each row in the table lists the name of the model the exercises where it is described a short description of the model and a URL where it can be found
162. nload free ebooks at bookboon com 155 Please click the advert Artificial Intelligence Exercises Agent Behaviour 10 Intelligence THINGS TO TRY Try out each of the solutions in turn to see what happens Which one seems to be closest to what happens in real life Try changing the values of the sliders to see what effect they have on each solution for turtle agents solutions 1 and 2 patch agents solution 1 boids agents solution landparticle agents solution 1 When the solution variable is set to particle agents 1 try sliding the max particles slider back and forth The effect of this is that the tap will seem to be temporarily turned on and off g and automatic nomical to create rience expertise and creativity ost performance beyond expectations eed the best employees who can challenge The Power of Knowledge Engineering q A Plug into The Power of Knowledge Engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 156 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence EXTENDING THE MODEL Try devising your own solution that simulates water flow more accurately Observe the process of problem solving that you apply in order to come up with your solution Would it be possible to devise a computer program to replicate your behaviour Exercise 10 8 2 Run the Water Flowing Uphill NetLogo model and set the solution choos
163. nt 03 n adaptive red Agent 01 Entropy Agent 01 Code Length 5 9 1 29 606 13 3 5 adaptive blue Agent 02 Entropy Agent 02 Code Length 4 296 1 53 785 5 o adaptive pink Agent 03 Entropy Agent 03 Code Length 4 1 25 633 36 Agent 02 s distribution Figure 7 8 1 Screenshot of the Interface for the Cars Guessing Game model after the setup button has been pressed followed by the animate cars button with the slider values as shown in the image WHAT IS IT This model plays a simple game trying to guess the colour of cars as they drive past Various agents maintain probability distributions that they use to predict the arrival of the cars The source distribution is a fixed distribution that is used to generate the cars so is the most accurate and therefore its entropy is the source entropy and corresponding code length is the lowest the code length is the optimal cost of encoding the sequence of cars given its probability distribution Agents 01 to 03 maintain fixed distributions which the user can adjust as they see fit by changing the slider values to the left of the Interface An adaptive agent also maintains a dynamic distribution by updating counts of cars that have previously been seen These counts are shown in the Interface by monitors under the heading Adaptive Agent s distribution The entropy and code length calculations are shown by monitors at the middle top of the Interface The adaptive entrop
164. nt of the door and have no key then try to open it R3 If the door does not open and I have no key then go to look for the key R4 IfI am looking for a key and there is a key in front of me then take the key and go towards the door Solution to Exercise 6 3 1 All the four models exhibit reactive behaviour because they all react directly to sensory input None of the models exhibit cognitive behaviour as none of them demonstrate higher level cognitive abilities such as the ability to acquire and process knowledge the ability to reason the ability to learn the ability to plan the ability to make rational decisions and so on All of the methods exhibit some form of self organisation and therefore exhibit an emergent behaviour as a result The turtles in the Flocking model organize themselves into flocks the turtles in the Fireflies model synchronise their flashing and the turtles in the Termites and State Machine Example models end up with Download free ebooks at bookboon com 164 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises woodchips gathered into piles This is done through local interactions rather than being imposed externally Solution to Exercise 6 3 2 The ANZ Continental Drift model illustrates one solution of how to get Australia and New Zealand back together again ANZ Continental Drift http files bookboon com ai ANZ Continental Drift
165. nt to the other two Try changing the chatbot mid conversation Clearly the conversational abilities of all three chatbots are poor But can you devise a conversation that seems believable Now what made it believable Can you change the chatbots to do this automatically EXTENDING THE MODEL Try adding your own rules to the chatbots You may need to turn the switch debug conversation to On to find out why some of your new rules do not work Remember the order the rules are processed is very important an earlier rule that matches will supercede a latter one resulting in the latter one never coming into effect So be very careful about where you place your rule Can you think of a way to avoid this problem NETLOGO FEATURES The model demonstrates the use of an extension where the code is written in Java Download free ebooks at bookboon com 151 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence Exercise 10 7 2 Describe how you might go about building a conversational agent to pass the Turing Test What are the major hurdles that need to be overcome Do you think that an Eliza like approach based on using regular expressions adequate for tackling this problem 10 8 Towards computers with problem solving ability Exercise 10 8 1 Water Flowing Uphill Model Try out the Water Flowing Uphill model in NetLogo MS ticks 0 setup environment go simulation Used by turtle agents so
166. ntation of the real life garden maze at Hampton Court Palace in the U K and the Chevening House maze which is a representation of the real life maze at Chevening House in the U K designed to thwart the hand on the wall method of getting to the centre of the maze The entrance of the empty maze is at the bottom middle of the maze and the goal is to reach the exit at the top of the maze The entrance of the Hampton Court maze is in the middle bottom and the goal is to reach the centre Similarly the entrance for the Chevening House maze is in the middle bottom and the goal is to reach the centre The problem is the following How do you reach the goal from the entrance To solve this problem search needs to be employed to try out the different paths that are possible This model implements some of the classic search algorithms and these can be applied to this toy problem to see how the different search strategies perform against each other WHAT IS ITS PURPOSE The purpose of this model is to show how some of the classic search algorithms such as breadth first search and depth first search can be implemented in NetLogo and also how they can then be applied to solving a classic toy problem in order to compare how they perform HOW IT WORKS The model implements the search algorithms in a novel way by making use of an agent oriented approach which is relatively easy to implement in NetLogo Rather than use a queue data structure to implement
167. ntic networks select problem This selects the type of problem Zoo Animals Identify the name of a zoo animal from observations New Zealand birds Guess the name of a New Zealand bird Sailing boats Identify the type of boat that is sailing past in a tall ships parade layout type This selects the type of layout to be used for visualising the states and paths in the event map spring length spring constant repulsion constant This can be used to modify the layout if the Layout type is spring trace on If set to On this will write output into the Output box that traces the reasoning process Output box This is used for writing the output of the reasoning process and other information into THINGS TO NOTICE Notice which of the types of knowledge representation make the reasoning process more difficult to understand when they are visualised using the event map format in the environment for example the walker agents for semantic networks seem to jump around a lot Notice which of the types of knowledge representation produce more complicated event maps for the different problems Download free ebooks at bookboon com 140 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge THINGS TO TRY Try the different types of knowledge representation for the three types of problems Try to see how they relate to each other and how they differ Find out what happens d
168. oad a language model first select which text you wish to load from the which text chooser then press the load model from text button This will load the language model into the environment so that it can be visualised using the layout specified by the layout type chooser If you have selected a Download free ebooks at bookboon com 72 Artificial Intelligence Exercises Agent Behaviour 7 Communication spring layout type then pressing the change layout button can sometimes remove some of the clutter once it has been loaded Other language models from different texts can be subsequently loaded if needed The pick these streams and pick these events input boxes can be used to filter out the language model so that only the part of the language model that matches the specific streams and events are visualised The text event chooser specifies what type of event should be loaded The NetLogo model at the moment only processes character and word events To see how the language model s process a testing text sequence first edit the test text input box and insert the test text sequence that you would like to be processed then press the go walkers button The paths taken through the language model s will then be highlighted as the testing text sequence is processed sequentially THE INTERFACE The NetLogo model s Interface buttons are defined as follows setup This re initialises all the variables change layout This cha
169. occur concurrently on separate streams The knowledge for three toy problems is implemented in the model These three problems are Zoo animals Identify the name of a zoo animal from observations New Zealand birds Guess the name of a New Zealand bird Sailing boats Identify the type of boat that is sailing past in a tall ships parade WHAT IS ITS PURPOSE The purpose of this model is to show how to represent knowledge using the traditional forms of knowledge representation such as first order logic frames rules and decision trees A secondary Download free ebooks at bookboon com 137 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge purpose is to show how these different forms of knowledge representation can be re cast and visualised within a new type knowledge of representation called event maps The event maps representation also facilitates the visualisation of the reasoning process as well HOW IT WORKS The different forms of knowledge are represented using state turtle agents and path link agents that are used to describe a transition path between the states State agents own three variables depth the depth in the event tree where the root of the tree is at depth 0 a child of the root state is at depth 1 and so on stream This is the name of the stream with which the event is associated event This is the event itself Events can be sensory events moto
170. oceeds Which search behaviour seems to be the best at this problem Which seems to be the worst Try out each of the different search behaviours to see which one is the most effective Find out how well each search algorithm performs against the four evaluation criteria time complexity space complexity optimality and complexity Relate the theory to what happens in practice Try switching the searching behaviour mid stream The difference in behaviours is most noticeable in the empty maze For example try switching from a breadth first search to a depth first search then back again after setting the move forward behaviour and move forward after turning behaviour chooser settings to Move forward n steps unless hit wall and set the move forward step amount slider to 1 ate to mainte thanks to our g and automatic jomical to create st performance beyond expectations ve need the best employees who can allenge Plug into The Power of Knowledge Engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 89 Artificial Intelligence Exercises Agent Behaviour 8 Search Try sliding the move forward step amount slider back and forth as the search proceeds if you have set either of the move forward behaviour and move forward after turning behaviour choosers to be Move forward n steps unless hit wall Try out the different heuristics for the informed searches Greedy Best Firs
171. ogo Download free ebooks at bookboon com 70 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication setup change layout 2 go walkers g kill all walkers load model from remove model layout type spring y which text English Bible Y td text event Chars y T max tree depth 2 i max text size 50 _Window size 40 1 _Spring cons 0 2 i Lspring enath 57 repulsion co 50 MAS ticks 0 ar mo HCY wm rE s T char ak et tip O Ae char b ne char Y B nist ar han f pierze B car har 0P chay g PP char G np char n 3P char a py model Eng char o 1P Char a a 1p D cha gait har h 1 char P pe char char e uP char v 1P cha 1P char n 17 bar d 1P ar h BP char dit e char are d nP char a BY char it cj gps IP pick these streams i Change pick these events Change test text Change In the beginning God cree ven and the earth Total of paths traver 6740 E o 0 Number of eve 58 8 Most Active Recently in Window English Bible Chars i on setup char models gn show counts On 2 ice debug trace Figure 7 9 1 Screenshot of the Interface for the Language Modelling model after the setup button has been pressed followed by the load mod
172. okboon com 48 Artificial Intelligence Exercises Agent Behaviour 7 Communication Exercise 7 3 2 What happens in the Language Change model when the agents communicate with each other In particular explain how the following procedures work in the model e communicate via e speak e listen 7 4 Communicating Behaviour Exercise 7 4 1 Communication T T Example 2 NetLogo Model Try out the Communication T T Example 2 model in NetLogo Communication For original model see NetLogo s Models Library Code Examples gt T T Example 2 Communication T T Example for modified model described here http files bookboon com ai Communication T T Example 2 nlogo 4b ticks 71 30 setup go y count turtles with color red 1000 number of people 1000 Percentage of people knowin o ofpopulation S Ticks 73 8 Figure 7 4 1 Screenshot of the Interface for the Communication T T Example 2 model after the setup button has been pressed followed by the go button with the number of people set at 1000 Download free ebooks at bookboon com 49 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication Read the What is it section in the Information for the original model Then try out the modified model whose modified Information is detailed below HOW TO USE IT Select the number of people you want in the simulation u
173. on New in the BehaviorSpace dialog box that appears This will create a further Experiment dialog box where you specify the settings for your experiment Enter a name for the experiment at the top of the dialog and use the values listed below for the remaining fields Vary variables as follows note brackets and quotation marks links per node 5 Layout type circle animate dijkstra true nodes to delete 0 centre of the universe Kevin Bacon new walkers per node 1 Ssimulation ticks 500 new walkers per tick 5 percent in both 10 network update 0 network type P2P no supernodes time to live 5 10 15 20 nodes to add 0 no of nodes 1000 revisits allowed false links per super node 20 no of super nodes 5 knowledge sharing None Word of mouth Blackboard Combined 1 Combined 2 Repetitions 5 Measure runs using these reporters report success rate Measure runs at every step y Setup commands setup network Go commands go walkers Stop condition Final commands Time limit 0 Download free ebooks at bookboon com 63 Artificial Intelligence Exercises Agent Behaviour 7 Communication 3 The field that begins Vary variables as follows sets the range of values for the possible settings for the model The Repetitions field specifies how many times the experiment is repeated The Measure runs using these reporters fie
174. onitor reports the number of nodes in the network It will change if the network is updated when the network update nodes to add and nodes to delete sliders are all non zero Walkers This monitor reports the number of active walker agents in the network Visited This monitor reports the number of nodes visited by the walker agents Success This monitor reports the percentage of walker agents that have been successful in reaching the goal node THINGS TO NOTICE Notice how well the agents perform at finding the goal node when the knowledge sharing method is set to None Compare this with the other knowledge sharing methods Notice that the blackboard knowledge sharing mechanism usually outperforms the word of mouth mechanism in terms of the percentage of successful searches Notice that pre loading the blackboards with shortest path information gained from execution of Dijkstra s algorithm can significantly improve the search The financial industry needs a strong software platform That s why we need you Find your next challenge at www simcorp com careers m www simcorp com S l m Co rp MITIGATE RISK REDUCE COST ENABLE GROWTH Download free ebooks at bookboon com 59 Artificial Intelligence Exercises Agent Behaviour 7 Communication Notice that when there is a learning effect due to the knowledge sharing mechanism the Successful Searches plot will start to curve upwards rather then continue as
175. orld BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizonsuniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 120 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge setup draw colour rectangle 2 draw colour circle Figure 9 4 3 Screenshot of the Interface for the Colour Cylinder model after the setup button has been pressed followed by the draw colour circle button WHAT IS IT This model demonstrates how colour can be represented in 3 dimensions hue saturation and brightness These correspond to the dimensions that experiments indicate humans use to perceive colour WHAT IS ITS PURPOSE The purpose of this model is to show that if humans rely on three dimensions to perceive colour hue saturation and brightness then it is not possible to exactly define precisely the concept of a particular colour such as red or yellow as it is not possible to define the boundaries between the different colours shown in the colour spindle or cylinder HOW IT WORKS A drawer agent is used to draw the colour circle or rectangle Two of the colou
176. ot layout type This specifies how the network should be laid out when it is visualised no of nodes This is the number of nodes to place in the network no of super nodes This is the number of super nodes to place in the network These are nodes that usually will have significantly more links than standard nodes as specified by links per super node slider links per node This specifies the maximum number of links a standard node will have The actual number chosen for a particular node will be a random number between and this number Download free ebooks at bookboon com 57 Artificial Intelligence Exercises Agent Behaviour links per super node This specifies the maximum number of links a super node will have The actual number chosen for a particular super node will be a random number between 1 and this number simulation ticks This specifies how long the simulation should run for network update If non zero this will cause the network to be updated at an interval according to a random number from 0 up to the value of this slider nodes to add The number of nodes specified by this slider will be added to the network when the next update occurs according to the net work update slider nodes to delete The number of nodes specified by this slider will be deleted from the network when the next update occurs according to the network update slider new walkers per tick This sets the number of walker ag
177. our The font size for all text is specified globally using the Settings button in the Interface The colour of the background for the map is specified using the background colour chooser The background is changed cleared using the clear background button An image can be used for the background instead of a single colour by pressing the import background button To save the entire map and all the settings to disk select the File menu from the NetLogo menu bar then select Export World option To load it back in again at a latter date use Import World instead Download free ebooks at bookboon com 128 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge THE INTERFACE The model s Interface buttons are defined as follows setup This clears the map and resets the background draw path This draws a path by having a turtle agent draw with the pen down between successive points as the mouse is dragged around The path colour is specified by the path colour chooser and the path size is specified using the path size slider end path This designates that the path is ended A subsequent press of the erase path button will not be successful in deleting previously drawn paths The only way to get rid of them is to draw over the top of them with another colour erase path This will erase the points one by one in a path that has just been drawn using the draw path button unless that p
178. paths will take longer to appear however Why EXTENDING THE MODEL Consider how to visualise the distance calculations in 3 or more dimensions RELATED MODELS See the Searching Mazes Missionaries and Cannibals and Searching for Kevin Bacon models to see where these distance metrics become useful as heuristic functions in various informed search algorithms Download free ebooks at bookboon com 112 Artificial Intelligence Exercises Agent Behaviour 8 Search Exercise 8 6 2 Show how the Greedy Best First Search and A Search algorithms search the network shown below Show the list of active searcher agents along with the nodes they are visiting The start node is node n1 and the star node is the goal node i e node nO Expand the nodes in the order of their node number For example expand node n1 before every other node expand node n2 before every other node except node n1 and so on Also allow nodes to be revisited How do these informed searches compare to the way uninformed search algorithms such as Breadth First Search and Depth First Search search the network See Exercise 8 3 1 8 7 Local search and optimisation Exercise 8 7 1 Show how the Hill Climbing Search searches the network shown below Show the list of active searcher agents along with the nodes they are visiting The start node is node n16 on the lower left corner and the star node is the goal node i e node n0 Expand the nodes in the order
179. piles Exercise 6 3 4 Is there some way of determining when an emergent property appears in a model 6 4 The Frame of Reference Problem Exercise 6 4 1 The default mode in NetLogo models is for the observer agent to be looking down on the world from above centered at the origin this is usually at the co ordinates 0 0 The frame of reference in this case is a third person perspective of the world That is the observer agent is external to the world being observed It is possible in NetLogo to change to a first person perspective using the follow ride and watch observer commands and the follow me ride me and watch me turtle commands Find out what each of these commands does Develop a model or models of your own to demonstrate their use Download free ebooks at bookboon com 17 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Exercise 6 4 2 The models in the Perspective Demos directory of the NetLogo s Models Library illustrate how these commands can be used to change the perspective from third person to first person for some of the models in the Models Library Have a play with the Ants Perspective Demo Flocking Perspective Demo and Termites Perspective Demo models Look at the code to see how the commands are used in the models Try out these models in both 2D and 3D Press the watch one of turtles and follow one of turtles buttons to see what happens How easy is it to discern the
180. ppens Select the different types of search by selecting the appropriate option using the search behaviour chooser For the informed searches Greedy Best First Search and A you will also be able to select from one of three heuristics Zero Euclidean Distance and Manhattan Distance with two further heuristics People on the left side and Min no boat trips needed available for the Missionaries and Cannibals model For the Searching Mazes model first set the maze being searched chooser in the Interface to Empty Maze Also find out what happens when you select the Hampton Court maze or the Chevening House maze Explain your results Exercise 8 8 2 Use the BehaviorSpace tool to find out which of the different search algorithms work best for the following NetLogo models Searching Mazes Missionaries and Cannibals and Searching for Kevin Bacon Do this in a more systematic manner than for Exercise 8 8 1 See Exercise 7 6 1 for more information on how to use the BehaviorSpace tool Download free ebooks at bookboon com 115 Artificial Intelligence Exercises Agent Behaviour 8 Search When filling in the Experiment dialog in the BehaviorSpace tool you can use the total searchers reporter for measuring the runs for all three models i e enter the text total searchers into the Measure runs using these reporters field You will also have to remove the various user messages in the
181. pply and for all current job openings please visit our web page www ericsson com careers 144 Download free ebooks at bookboon com Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Database ABCD Rules base Rule 1 IF X is true AND Y is true AND A is true THEN Z is true Rule 2 IF V is true AND B is true THEN Y is true Rule 3 IF V is true AND C is true THEN X is true Rule 4 IF D is true THEN V is true i Use forward chaining to determine if the fact Z can be inferred from this knowledge base Show the state of the database and the rule that was fired after each step in the reasoning process ii Use backward chaining on the same knowledge base to determine if the goal Z succeeds Show the state of the database and the rule that was matched after each step in the reasoning process 9 11 Knowledge and reasoning using frames Exercise 9 11 1 Represent the knowledge contained in each of the sentences for Exercise 9 9 1 using frames and compare with the solutions you devised for production rules and first order logic Which sentences do you have problems with when trying to represent the knowledge and why Exercise 9 11 2 Modify the Knowledge Representation model so that the frames knowledge is explicitly defined for the three problems implemented in the model instead of being converted from rules as is done in the convert rules to frames base
182. procedure You will need to modify the setup frames base procedure so that it is similar to the setup semantic network procedure and also define a setup frame procedure for defining a specific frame 9 12 Knowledge and reasoning using decision trees Exercise 9 12 1 Modify the Knowledge Representation model so that the decision tree knowledge is explicitly defined for the three problems implemented in the model instead of being converted from rules as is done in the convert rules to decision tree procedure You will need to define a Download free ebooks at bookboon com 145 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge setup decision forest procedure similar to the setup semantic network procedure and also define a setup decision tree procedure for defining a specific decision tree 9 13 Knowledge and reasoning using semantic networks Exercise 9 13 1 Redo Exercise 9 8 3 by running the Knowledge Representation model again for the three different problems as specified by the select problem chooser and the different types of knowledge representation as specified by the KR type chooser This time play close attention to how the knowledge is visualised in the environment for the different types and observe what happens during the reasoning process when the go reason button is pressed Which of the types seem easier to follow what is happening In particular note what happens wi
183. pturing the knowledge being expressed in the text Solution to Exercise 9 9 2 i dx Car x A Red x Some cars are red ii Vx Cake x gt Sweet x All cakes are sweet iii AxVy Person x A Cake y A Likes x y Not all people like cakes some people do not like cakes Solution to Exercise 9 9 3 i All cats like fish Vx Cat x gt Likes x Fish ii Bananas are either green or yellow or brown Vx Banana x gt Green x v Yellow x v Brown x iii Some people eat everything AxVy Person x A Eats x y Solution to Exercise 9 10 1 As a short hand notation we can convert the database and rules from the IF THEN format to the equivalent format shown below Database ABCD Rules base Rule 1 X amp Y amp A gt Z Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Answer for i Rules fired are marked by Soulutions to Selected Exercises Download free ebooks at bookboon com 194 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Cycle 1 Database ABCD Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Action Add V to database Cycle 2 Database ABCDYV Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rul 3 V amp C gt xX Rul 4 D gt V Le Le Action Add Y to database matter
184. r 9 Knowledge 9 8 Representing knowledge using event maps Exercise 9 8 1 Central Park NetLogo Model Try out the Central Park Events model in NetLogo Central Park Events http files bookboon com ai Central Park Events html M4 ticks 0 30 setup change layoutey eel spring con 0 35 eee spring length 20 repulsion co 29 Figure 9 8 1 Screenshot of the Interface for the Central Park Events model after the setup button has been pressed followed by the change layout button combined with moving the spring length slider down to 0 and back up to 20 WHAT IS IT This model visualises a small set of events that an agent needs to perform if they wish to go from the Zoo to the Boat Pond in New York s Central Park The events are shown using an event map representation where events are linked to other events on separate streams in an ordered sequence The idea is that an agent processes events simultaneously on separate streams The approach is similar to the approach adopted for Event Stream Processing ESP Download free ebooks at bookboon com 133 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge WHAT IS ITS PURPOSE The purpose of this model is to show how to visualise a series of events using an event map HOW IT WORKS The model uses turtle agents to represents the states in the event map and uses links agents to represent the paths bet
185. r dimensions hue and saturation determine what the colour circle looks like for a specific value of brightness Peter G rdenfors work with conceptual spaces theory talks about how humans perceive colour using a colour spindle where the radius of the circle gets progressively smaller as the brightness is increased or decreased eventually ending up at points for when the brightness is maximum or Download free ebooks at bookboon com 121 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge minimum In the model this is when the brightness is set to 255 and 0 and corresponds to the colours white and black respectively As the model s visualisation is limited to 2D the model depicts a colour cylinder instead of a colour spindle with each colour circle representing a slice of the cylinder for a specific brightness value HOW TO USE IT Press the setup button first Then press either the draw colour rectangle or draw colour circle button next The brightness slider determines the brightness of the colours that are drawn The hue increment and saturation increment sliders can be used to control how fine the colour is drawn THE INTERFACE The model s Interface buttons are defined as follows setup This clears everything makes the background white and creates a drawer agent ready to draw the colour circle or rectangle draw colour rectangle This draws a colour rectangle for a specific brightn
186. r events or abstract events the latter is sensory motor and abstract events that when they occur together cause this event to occur simultaneously as well The states and paths form a finite state network Walker turtle agents are used to walk the event map network during the reasoning process The paths i e transitions are not labelled they just indicate the order with which events occur The network of states and paths can be thought analogously of as a map of events hence why it is called an event map The event map may or may not be a faithful representation of what is really happening in the environment i e in this model the event map may or may not be a faithful representation of the knowledge that it is trying to represent Yow re full of energy and ideas And that s just what we are looking for UBS 2010 All rights reserved Looking for a career where your ideas could really make a difference UBS s Graduate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates www ubs com graduates a UB S Download free ebooks at bookboon com 138 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Event maps can be used to implement all the standard types
187. r hsb as rgb set this to a specific color if you want to draw a border to the circle Change hsb as rgb in the above code to the colour you want for your border For example if you want it to be black set it to 0 if you want it to be blue set it to 105 EXTENDING THE MODEL Have a go at adding labels to the colour circle such as red yellow and green Alternatively set up a HubNet activity where each student chooses where to place colour labels See how well they match across the classroom ience expertise and creativity erformance beyond expectations Plug into The Power of Knowledge engineering Visit us at www skf com knowledge Download free ebooks at bookboon com 123 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge NETLOGO FEATURES It uses the NetLogo command hsb to generate the patch colours based on the hue saturation and brightness values The patch at heading and distance command is used to draw the colour circle 9 5 Knowledge engineering problems Exercise 9 5 1 List two or more real life examples for the following types of problems for intelligent systems e classification e clustering e control e diagnosis e optimisation e prediction e selection How can knowledge be leveraged by an intelligent system for these types of problems Is the nature of these problems inherently different requiring that knowledge is processed in a fundamentally different way
188. r more information visit www london edu mim email mim london edu or call 44 0 20 7000 7573 Download free ebooks at bookboon com 135 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge EXTENDING THE MODEL Combine this model with the Map Drawing model loaded using the Map Drawing Central Park World Then show several agents wandering between the Zoo and Boat Pond by animating the map and event map at the same time This can be done so that each agent ends up with slight variations in the paths they choose since there is the possibility for minor variations to occur between the states of the event map For example when turning left the agent may choose to turn left by slightly more or slightly less than 90 degrees Similarly when walking forwards the agent can choose to walk forwards a distance that varies slightly to other agents applying the same event map NETLOGO FEATURES The model uses the layout spring command for modifying the layout of the network of states turtle agents and paths link agents RELATED MODELS See the Wall Following Events model and the Knowledge Representation model Exercise 9 8 2 Knowledge Representation NetLogo Model Try out the Knowledge Representation model in NetLogo Download free ebooks at bookboon com 136 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge setup knowledge dump knowledge reset reasoning RA o ch
189. r the environment and variables and re load the event map Normally this will appear in a cluttered form and the change layout button needs to be pressed subsequently change layout This can be used to clear some of the clutter by changing the values in the three sliders The model s Interface sliders are defined as follows spring constant This is a value used by the Layout spring command Changing it will usually not affect the visualisation of the event map much sprint length This modifies the length of the paths between the states of the event map network repulsion constant This controls how much each of the states repulse each other Download free ebooks at bookboon com 22 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THINGS TO NOTICE Notice how the repulsion constant slider can be used to repel the states away from each other for larger values and attract the states towards each other for smaller values Notice that the clutter in the network layout can often be removed by setting the value of the spring length slider to zero and then increasing it afterwards THINGS TO TRY Try altering the values of the sliders to see what effect this has on the layout EXTENDING THE MODEL Combine this model with the Wall Following Example 2 model Then animate the environment and event map at the same time for one of the agents in the env
190. raduate as an engineer or geoscientist Or have you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com 80 Artificial Intelligence Exercises Agent Behaviour 7 Communication In contrast notice how well the language model does at predicting the testing text if it is trained on exactly the same text i e the training and texts are exactly the same Note This situation seldomly appears in real life prediction because usually the testing sequence is completely unpredictable Note also that the best the language model for higher order models can do for predicting the next character is often only 1 2 or 1 bit to encode even in the situation where it has been trained on the testing text Why is this Hint What would happen if the length of the training text was much longer THINGS TO TRY Try altering the max depth of tree to see how this affects the prediction Try different combinations of training and testing texts EXTENDING THE MODEL Extend the PPMC language model so that it implements full exclusions and update exclusions Can you devise further methods to improve the prediction of the language model RELATED MODELS See the Language Modelling NetLogo model and the Cars Guessing Game NetLogo model 7 11 Communicating Meaning Exercise 7 11 1 Fill in the conceptual metaphor for the following two verses of
191. raw text over the top of the image at the point the mouse is next clicked Note that the right side of the text will be set to the mouse click point rather than the left The content of the text is specified by the text to insert input box To update this the user must click inside the input box then edit the existing text or clear it and replace it with new text The text update is only completed when the user types the return key The colour of the text is specified by the text colour slider draw mark This will draw a mark over the top of the image at the point the mouse is next clicked select object This will select an object if it is near to the next mouse click point The object will have a halo placed around it to show when it has been selected Subsequently dragging the mouse will result in the selected object being dragged around An object is an arrow head or tail a circle text or a mark When dragging the head or tail of an arrow the other end will remain stationary i e the arrow will pivot around the other end erase object This will erase an object if it is near to the next mouse click point If an arrow head or tail is erased then the shaft is erased as well and only the other end will remain select region This allows multiple objects to be selected Click using the mouse then drag A rectangular box will delimit the region of objects that are selected Temporarily stop clicking the mouse then click back somew
192. rawn as a dark brown curve from top left to bottom right The fire was ignited in the centre of the burnt out crimson circle and is shown to be expanding outwards on the outer rim of the circle except where it is blocked by the firebreak WHAT S MISSING IN THIS EQUATION Please click the advert MAERSK INTERNATIONAL TECHNOLOGY amp SCIENCE PROGRAMME Are you about to graduate as an engineer or geoscientist Or have you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com 14 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour Add a fourth button import forest for importing a forest and its surrounding geography directly from a satellite image or aerial photo so you can use the model to simulate how a fire might spread in a real life environment Use the extended model to investigate how the geography and different types of firebreaks affect the behaviour of the fire 6 2 Reactive versus Cognitive Agents Exercise 6 2 1 An agent wishes to go through a door but the door is locked An agent adopting the cognitive approach has constructed the following plan in its head Plan for opening the door Go to the place where the key is Take the key Go to the door Open the door with the key Suggest some rules that a reactive agent could use to solve the same task What are the lim
193. rgument clear al1l This clears all the variables setup regex This sets up the complete regular expression system and searches for the target specified as the first argument HOW TO USE IT First press the setup button in the Interface this will load the rules for each chatbot Then choose the chatbot you would like to chat with using the bot chooser either liza harry or both Then start chatting by pressing the chat button Download free ebooks at bookboon com 150 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence THE INTERFACE The Interface buttons are defined as follows setup This loads the rules for each chatbots chat This starts or continues the conversation with the chatbot that was selected using the bot chooser The Interface chooser and switch is defined as follows bot This sets the chatbot to the Liza chatbot the Harry chatbot or Both debug conversation If this is set to On debug information is also printed showing which rules matched THINGS TO NOTICE Harry seems to do a bit better at being paranoid than Liza does at being a Rogerian psychotherapist The latter also tends to repeat things a lot However it doesn t take long for the conversation with both chatbots to break down THINGS TO TRY Try out the different chatbots by changing the bot chooser Does Liza seem to have a different personality to Harry or is Both any differe
194. rrow turtle agent is given a circle shape and the end arrow turtle agent is drawn using NetLogo s default arrowhead shape The shaft of the arrow is drawn by creating a link between the two arrow turtle agents Circles and marks are standard turtle agents with a specific size colour and thickness drawn at a specific location using the mouse Text is drawn by setting the label of turtle agents that are drawn with size 0 i e only the label of the agent is made visible not the agent itself NetLogo s distancxy reporter is used to find the nearest object by the select object and erase object procedures The watch command is used to highlight the selected object The global variable selected is used by the select region and erase region procedures to hold an agentset of objects that are within the specified region The entire agentset can then easily be dragged together or erased in unison using a single ask command HOW TO USE IT Pressing setup clears the environment and then loads the image as specified by the map or image to use chooser Then the user can place arrows circles text and marks over the top of the image using the draw arrow draw circle draw text and draw mark buttons by pointing then clicking using the mouse Download free ebooks at bookboon com 188 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises THE INTERFACE The buttons in the Interface are
195. s purpose is to show how entropy and code length calculations are made given a probability distribution Central Park Events 9 8 1 This model visualises a sequence of events that are necessary for going from the Zoo to the Boat Pond in Central Park New York i Being Kevin Bacon 7 5 1 This model implements various algorithms related to communication amongst agents in a network such as Dijkstra s algorithm and communication via word of mouth or using blackboards It also demonstrates some important concepts such as the small world phenomenon degrees of separation and super nodes in peer to peer networks Chatbot 10 7 1 This model implements two basic chatbots Liza and Harry using regular expressions Colour Cylinder 9 4 3 This model demonstrates how colour can be represented in 3 dimensions hue saturation and brightness l t Communication T T TED 2 7 m D This model simulates the spreading of a message between agents This is a modification of the Communication T T Example model in NetLogo s Models Library Code Examples gt Communication T T Example for modified model used here Crowd Path Following 6 7 1 This model is an attempt to recreate boids ine cabs pee AURI that use the crowd path following steering behaviour htt es book ath Fo t Entropy Calculator 7 7 1 Solution to 7 7 D This model allows the user to calculate the entropy for a specific probability distribution
196. s shown with a halo Make sure to click on the centre of the object otherwise a halo showing it has been selected will not appear select region This allows the user to select a rectangular region of objects for moving around The user clicks and drags the mouse to define the region being selected this is shown with a red box dynamically as the mouse is dragged all objects that can be moved in the region are also highlighted To move the selected objects around lift the mouse momentarily when the rectangular region is as you want it click back inside the red rectangle then drag it to the desired location erase region This will erase all the objects that have been selected using the select region button clear background This clears the existing background and redraws it using the colour specified by the background colour chooser import background This allows the user to use an image file as the background draw patches path This draws a path using patch agents The size of the patches is specified by the patches path size slider draw turtles path This will draw a path using turtle agents The size of the turtles is specified by the turtles path size slider draw text This draws the text specified in the text Input box on the map The right side of the text ends up at the point the mouse is clicked The colour of the text is specified by the text colour chooser The font size for all text can be changed globally
197. sing the number of people slider press the setup button then press the go button THE INTERFACE The buttons in the Interface are defined as follows setup This clears the variables and resets the simulation creating persons at random locations whose number is specified by the number of people slider go This starts the simulation The monitor slider and plot in the Interface are defined as follows Download free ebooks at bookboon com 50 Artificial Intelligence Exercises Agent Behaviour 7 Communication count turtles with color red This counts the number of turtles who know the message number of people This is the number of persons who will be created next time the setup button is pressed Percentage of people knowing message This plots the percentage of the people who know the message against the number of ticks THINGS TO NOTICE Notice that when the number of people is set to maximum the message literally spreads like fire THINGS TO TRY Try running the simulation with different number of people Notice what happens in the plot when the number of people is small rather than large It takes a lot longer to spread the message when the population size is small Why is this since the number of people do not know the message at the beginning is a lot less EXTENDING THE MODEL Try changing the model so multiple messages are being spread Alternatively try having two different breeds of
198. suniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 86 Artificial Intelligence Exercises Agent Behaviour 8 Search HOW TO USE IT To initialise the search and draw the selected maze specified by the maze being searched chooser press the setup button in the Interface A turtle agent with a shape of a person will slowly walk towards the entrance of the maze Then the entrance will be blocked to prevent the agent from exiting in that direction To start the search either press the go once button to make the search proceed one step at a time or press the go forever button to make the search proceed continuously until it reaches the goal state or it gets stuck THE INTERFACE The model s Interface buttons are defined as follows setup This will clear the environment and all variables and redraw the selected maze chosen by using the maze being searched slider go once This will make the search proceed one step at a time go forever This will make the proceed continuously until it has reached the goal state or it gets stuck The model s Interface choosers sliders and switch are defined as follows maze being searched This selects the maze to be searched There are three available the empty maze the Hampton Court maze and the Chevening House maze search behaviour This specifies what search strategy the agents should employ when performing
199. swer our questions in a knowledgeable manner In this definition knowledge is explicitly associated with some agent and question answering is a key part to being knowledgeable With this definition a knowledge based system must act like a knowledgeable agent Humans judge the agent to be knowledgeable or not as the case may be i e A secondary agent is doing the testing similar to the Turing Test Just as importantly it is the interaction between the testing agent and the possibly knowledgeable agent that is the most important part of the process This is a much stricter definition of what a knowledge based system is and in this light a traditional logic based definition which states that rules of inference are the important distinction to distinguish between a knowledge base and database seems inadequate as it requires someone to know the meaning of the output of the inference process i e what it all means and how we can apply the knowledge gained through the inference Note In expert systems quite often the knowledge base is defined as being the facts plus rules plus inferencing system whereas the database is simply the facts However with this definition can the system justifiably be called a repository of knowledge in the common meaning of the term Download free ebooks at bookboon com 184 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to 9
200. t Search and A Search Do they have any effect on how effective the search is Try out the all the different combinations of the move forward behaviour and move forward after turning behaviour chooser settings and the move forward step amount slider setting This results in a huge variety of behaviours for the searching turtle agents For example have a go at reproducing the behaviour shown by the screenshots in Chapter 8 e g Figure 8 10 of the book Artificial Intelligence Agent Behaviour I Try changing the parameters of each maze by altering the settings of the sliders on the right EXTENDING THE MODEL Try adding other search algorithms or adding your own variations to the ones implemented in the model Try adding your own maze perhaps of a garden maze or other maze that exists in real life If your maze does have a real life equivalent note the discrepancies that you have to introduce into the virtual representation where it does not exactly correspond to what is happening in the real life maze NETLOGO FEATURES Note the use of the hatch and die commands to clone child searchers and terminate parent searchers thus avoiding the need for a separate queue data structure to maintain the information during the search Note also the variation in the ask command that determines the strategy being used that defines the different searches RELATED MODELS For other search problems see the Searching For Kevin Bacon model and the Se
201. t a letter date press the Load Network from a File button To generate a problem description that could be used for an exam or test question for example then press the Generate Problem Description button This will write the description into the Output box shown on the middle right bottom of the Interface To add further nodes to the network press the Add to Network button THE INTERFACE The model s Interface buttons are defined as follows Generate Problem Description This will generate a description in English of the search problem and write it into the Output box The problem description is based on the values of the choosers and sliders in the Interface It can be used for setting an exam or test question or set an exercise for an assignment as course work Add to Network This will allow the user to add nodes to the network including networks that have been generated randomly Load Network from a File This will load a network layout that has been previously saved to a file on disk Save Network to a File This will save the network layout to a file on disk This can be loaded at a latter date using the Load Network from a File button The model s Interface sliders are defined as follows link thickness This sets the thickness of the links that are drawn between the nodes in the network spring constant spring length repulsion constant These sliders are used by the change layout algorithm when the layout
202. t attempt to maintain an optimum temperature around themselves In NetLogo s Models Library Biology gt Heatbugs http ccl northwestern edu netlogo models Heatbugs Knowledge Representation 9 8 2 This model visualises the knowledge and reasoning processes for three toy problems using different methods for knowledge representation http files bookboon com ai Knowledge Representation html Language Change 7 3 1 This model shows how social networks amongst language users can affect the course of language change In NetLogo Model s Library Social Science gt Language Change http ccl northwestern edu netlogo models LanguageChange ericsson com YOUR CHANCE TO CHANGE THE WORLD Here at Ericsson we have a deep rooted belief that the innovations we make on a daily basis can have a profound effect on making the world a better place for people business and society Join us In Germany we are especially looking for graduates as Integration Engineers for e Radio Access and IP Networks e IMS and IPTV We are looking forward to getting your application To apply and for all current job openings please visit our web page www ericsson com careers Download free ebooks at bookboon com 10 Artificial Intelligence Exercises Agent Behaviour Preface Language Modelling 7 9 1 This model shows how a language model can be constructed from some training text Its purpose is to s
203. t bookboon com 83 Artificial Intelligence Exercises Agent Behaviour 8 Search 8 Search 8 1 Search Behaviour Exercise 8 1 1 Research in the literature how the terms search and research are being defined Observe the methods you use to perform the research Are there distinctly different meanings of these terms or do they all spring from the same generic meaning 8 2 Search Problems Exercise 8 2 1 Searching Mazes NetLogo Model Try out the Searching Mazes model in NetLogo Empty Maze parameters Hampton Court and Chevening House Mazes parameters T T oi a pop rok i ae i i ill ti Active Searcher Agents j 215 46 ms i Maximum Active Searcher Agents ia 5 an a 5 H Total Searcher Agents i 1684 T For multi agent depth first search only 0 For depth limited and Figure 8 2 1 Screenshot of the Interface for the Searching Mazes model for the Chevening House Maze after the setup button has been pressed followed by the go forever button with the slider values as shown in the image Download free ebooks at bookboon com 84 Artificial Intelligence Exercises Agent Behaviour 8 Search WHAT IS IT This model applies standard search algorithms to the problem of searching mazes Three mazes are provided in the model the empty maze with no internal walls the Hampton Court maze which is a schematic represe
204. t by professors and professionals from all over the world BBA in Global Business MBA in International Management International Marketing DBA in International Business International Management MA in International Education MA in Cross Cultural Communication MA in Foreign Languages Innovative Practical Flexible Affordable Visit www HorizonsUniversity org Write Admissions horizonsuniversity org Call 01 42 77 20 66 www HorizonsUniversity org Download free ebooks at bookboon com 53 Artificial Intelligence Exercises Agent Behaviour 7 Communication HOW IT WORKS The model implements the search algorithms in a novel way by making use of an agent oriented approach which is relatively easy to implement in NetLogo The model adopts a purely agent oriented solution where the information is distributed amongst NetLogo agents Unlike the traditional search algorithms such as uninformed searches like Breadth first Search and Depth first Search or the informed searches such as Greedy Best first Search and A that are implemented in the Searching for Kevin Bacon model these agents conduct the search by sharing knowledge with other agents that are concurrently conducting the search In the word of mouth knowledge sharing method if agents have found the goal then they share their knowledge with other agents they meet while returning to their start destination With the blackboard method the agents share th
205. t comes with NetLogo s models library It simulates the spread of a fire through a forest This model allows the user to add firebreaks extra forest and ignition points It also allows the user to import a small area taken from a simulated natural colour satellite image of Chateaubriant France in order to test out what might happen in a real life environment Download free ebooks at bookboon com 158 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises HOW IT WORKS The setup button will set up a random forest with the specified density If the user presses the import forest button then the satellite image of Chateaubriant France is imported instead The user can start fires at various points by pressing the ignite forest button and then clicking the mouse button at the required location Alternatively the user can choose to ignite the fire on a continuous vertical line at the left of the environment using the ignite on the left button Extra forest can be drawn using the draw forest button and firebreaks can be drawn using the draw firebreak button Pressing the go button will then start the simulation and the fire will then spread to neighbouring trees in four directions N S E and W from the initial ignition points The spread of the fire depends on the density of the surrounding forest There is no wind being simulated in the model so the fire can only spread to a
206. t moves forward each tick rate of random turn This controls how much the wandering boid turns each time tick The boid has a tendency to head in a right turning direction as the rate of random turn to the right as specified by the slider is twice that of the rate of random turn to the left vradius angle This defines the radius angle of the boid s vision cone radius length This defines the radius length of the boid s vision cone HOW TO USE IT Press the Set up button first then press Go You can draw extra obstacles by pressing the Draw Obstacle button and by then holding down the mouse at the point where you want the obstacles to be drawn You can change the frame of reference so that the visualisation is centred around where the boid currently is situated by pressing the Follow Wanderer button Download free ebooks at bookboon com 37 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour THINGS TO NOTICE Setting the boid s vision cone radius length or radius angle will result in the boid not seeing the obstacles and as a result will run right over the top of them Why What needs to be fixed so that even without any sensing ability it will still bounce off the obstacles Increasing the radius length while keeping the other variables the same discernibly changes the behaviour of the boid Try doing this dynamically while moving the radius length slider back and for
207. t the primary purpose of the model HOW TO USE IT To reset the simulation and load the present day shapes and relative positions of Australia and New Zealand press the setup button To start the process of winding the clock back so that New Zealand gradually creeps it way back to Australia press the continental drift button If you want New Zealand to end up at a different location then change the values of the sliders to suit THE INTERFACE The buttons in the Interface are defined as follows setup This button clears the environment and resets the variables then loads the map of present day Australia and New Zealand continental drift This button simulates the process of continental drift by drawing the relative positions of the two countries as the clock is wound backwards The sliders and monitor in the Interface are defined as follows max ticks This is the maximum number of ticks that the model should be run for If this number is too high then New Zealand will eventually run through the middle of Australia Download free ebooks at bookboon com 167 Please click the advert Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises unless the values of the other sliders means that it ends up heading elsewhere If this number is too low then the animation will stop early years per tick This sets the number of years per tick of the simulation This number multip
208. th Instead of covering most of the environment when the radius length is large then the boid covers a much smaller more constrained area usually at the top of the environment Sometimes it can get stuck seemingly trapped in the same place THINGS TO TRY Try adjusting the boid s speed radius angle and radius length to see how this affects the boid s behaviour Also try changing the Interface Settings to see if this has any affect Try adding obstacles to see how this affects the boid s ability to cover the entire environment For example add obstacles in the form of a maze Try to create black spots where the boid never visits Alternatively try to trap the boid into a small area Yow re full of energy and ideas And that s just what we are looking for UBS 2010 All rights reserved Looking for a career where your ideas could really make a difference UBS s Graduate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates www ubs com graduates a UB S Download free ebooks at bookboon com 38 Artificial Intelligence Exercises Agent Behaviour 6 Behaviour EXTENDING THE MODEL The model could be extended to add gradual acceleration and deceleration This would
209. th the semantic networks type of reasoning Is there a way of improving the visualization of the reasoning process of semantic networks or is this an inherent problem with this type of knowledge representation WHAT S MISSING IN THIS EQUATION You could be one of our future talents 7 5 A wo MAERSK INTERNATIONAL TECHNOLOGY amp SCIENCE PROGRAMME Are you about to graduate as an engineer or geoscientist Or have you already graduated If so there may be an exciting future for you with A P Moller Maersk MAERSK www maersk com mitas Download free ebooks at bookboon com 146 Artificial Intelligence Exercises Agent Behaviour 10 Intelligence 10 Intelligence 10 1 The nature of intelligence Exercise 10 1 1 Conversational ability and problem solving ability are postulated as two important ingredients for human level intelligence in this chapter Suggest some further ingredients and justify why you think they are necessary for your recipe for intelligence 10 2 Intelligence without representation and reason Exercise 10 2 1 What do we mean by representation and reason when we talk about intelligence without representation and reason Is it fair to say that Brooks insect like robots and the ant and termite agents in the Ants and Termites models do not represent knowledge and reason in some manner After all the intelligence is represented globally
210. the beginning obstacle colour This is the colour used for drawing the obstacles using the Draw button speed This controls the speed of the bird agents vision This defines the radius of the bird agent s cone of vision The angle is 360 degrees so the bird can see in all directions minimum separation This defines the minimum separation for the flockmates before the flockmates will start to separate Download free ebooks at bookboon com 31 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour max align turn max cohere turn max separate turn These sliders control the maximum angle a bird can turn as a result of each rule alignment cohesion and separation respectively The choosers in the Interface are defined as follows Filename This is the name of the image file which is loaded into the background when the Load Image button is pressed patch name This is the name of the image file which is converted into patches in the environment when the Load Patches button is pressed THINGS TO NOTICE Around obstacles sometimes the flock splits in two to head around different sides of an obstacle but then returns to a single flock The same behaviour can be seen in Craig Reynold s simulations What are the conditions that flock splitting seems to occur in Does it sometimes occur that the flock remains split in two Try this HF Ki options
211. the classic search algorithms this is a standard solution see for example Russell amp Norvig s AI textbook the model instead adopts a purely agent oriented solution where the information is distributed amongst NetLogo agents These agents conduct the search by transferring information to other agents that continue the search Hence an explicit queue data structure separate from what is happening in the search is not needed although technically the information is still being stored in a queue behind the scenes but the implementation of this is hidden behind the top level commands that NetLogo provides The hope is that this queueless agent oriented approach provides a more intuitive solution that makes it easier to grasp how the search strategies work It uses a situated embodied first person design perspective therefore arguably making it easier to understand the essential differences between the search strategies Download free ebooks at bookboon com 85 Please click the advert Artificial ntelligence Exercises Agent Behaviour 8 Search The breed of searcher turtle agents is used for the searchers that move throughout the maze trying to get to the goal These searcher agents maintain information about the current state of the search such as the time taken and the path and estimated costs Each searcher agent expands the search in three directions forward a right 90 degrees turn then forward and
212. the search A full description of the different search strategies can be found in Chapter 8 of the book Artificial Intelligence Agent Behaviour I heuristic This specifies the heuristic to be used for the Informed searches Greedy Best First Search and A Search move forward behaviour This specifies the type of move forward behaviour the searcher agent executes For an explanation of the options see the How It Works section above Download free ebooks at bookboon com 87 Artificial Intelligence Exercises Agent Behaviour 8 Search move forward behaviour after turning This specifies the type of move forward behaviour the searcher agent executes after it has turned either right or left For an explanation of the options see the How It Works section above move forward step amount This is the amount of steps forward the agent makes when it moves forward either directly ahead of itself on its current heading or after it has made a right or left turn allow revisited states This is a flag that if set to On will allow the search to proceed to already visited states left cols right cols above rows above cols entrance cols These sliders set the height and width of the empty maze and the size of the entrance col patches width row patches width These sliders set the height and width of the Hampton Court and Chevening House mazes max agents to expand This sets the maximum number o
213. tificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises NETLOGO FEATURES The import pcolors command is used to load the map of Australia and New Zealand The stamp command is used to stamp the shape of New Zealand at various stages of the simulation Note also the use of the sprout command to create a turtle agent for each patch agent where the colour is changed to black for New Zealand and yellow and green for Australia The change of colour makes the coding much easier for identifying which agents to shift around Either changing part of the environment for example by adding a different coloured border around the outside or changing a specific set of agents by altering a property such as colour is a common trick that can be used in NetLogo to greatly simplify the coding Solution to Exercise 6 3 3 The State Machine Example 2 model provides one solution to this problem State Machine Example 2 http files bookboon com ai State Machine Example 2 html number F density i setup go y pileness measure Number of empty spaces v pileness 16832 pileness plot N gt o pileness 0 Figure S 6 3 3 Screenshot of the Interface for the State Machine Example 2 model after the setup button has been pressed followed by the go button with the slider values as shown in the image Ab ticks 34614 35 1
214. till spreading Try using the BehaviorSpace feature see Tools menu to see the affect on the percentage of trees burnt for different slider values See Exercise 7 6 1 for more information on how to use the BehaviorSpace tool Try changing the way the fire spreads using the neighbours reporter instead of the neighbors 4 reporter in the code How does this affect the ability of the fire to spread EXTENDING THE MODEL Add wind to the model so that the fire has a chance to jump over firebreaks Add two types of people agents to the model firstly fire fighters who autonomously fight the fire by creating firebreaks and using water to put out the fire and secondly fire lighters who run around lighting the fire Add a further helicopter breed of agent that will fly over the forest with a large water bucket trying to put out the fire Add the ability to import any image file not just the one used by the model RELATED MODELS See the implementation of Breadth First Search in the Searching Mazes Missionaries and Cannibals and Searching for Kevin Bacon models CREDITS AND REFERENCES The particle agents code used in this model was obtained from the Particle System Waterfall produced by Uri Wilensky See the model for a list of references and the How to Cite section below HOW TO CITE If you mention this model in an academic publication we ask that you include citations for the original Fire model and for the NetLogo sof
215. tretched as in the model or an added rotation is required to rotate the country slightly around a central rotation point such as the present day location of Picton while New Zealand is being translated in a north west direction towards Australia THINGS TO TRY Play around with different values in the sliders Try to get New Zealand to miss Australia all together or alternatively have New Zealand end up wafting through the middle of it Try changing the values of the sliders during the middle of the simulation to see if you can change New Zealand s direction dynamically Especially try dynamically changing the value of the NZ xcor warp factor and NZ ycor warp factor sliders Warping New Zealand will change its direction slightly Why This can also be changed back again at a latter stage of the simulation EXTENDING THE MODEL Change the model so that Australia moves as well This will make the animation look better but will also slow it down Why Have a go at changing the simulation in the model so that that it more accurately reflects projections that are based on scientific evidence Once you have perfected it for the relatively straightforward situation of the drifting apart of Australia and New Zealand then adapt your model so that it works for the entire world Run your model backwards until you reach the time when there was just a single super continent Pangaea Download free ebooks at bookboon com 169 Ar
216. ts the entropy and code lengths for these distributions Fast track your career Masters in Management Stand out from the crowd Designed for graduates with less than one year of full time postgraduate work experience London Business School s Masters in Management will expand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School Regent s Park London NW1 4SA Applications are now open for entry in September 2011 United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Download free ebooks at bookboon com 68 Artificial Intelligence Exercises Agent Behaviour 7 Communication THE INTERFACE The model s Interface buttons are defined as follows setup This resets the animation and initialises the counts entropy and code length values animate cars This starts the animation in the environment the light blue rectangle shown middle bottom of the Interface The model s Interface sliders have the following naming convention lt agent name
217. tware Download free ebooks at bookboon com 163 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Wilensky U 1997 NetLogo Fire model _http ccl northwestern edu netlogo models Fire Center for Connected Learning and Computer Based Modeling Northwestern University Evanston IL Wilensky U 1999 NetLogo http ccl northwestern edu netlogo Center for Connected Learning and Computer Based Modeling Northwestern University Evanston IL In other publications please use Copyright 1997 Uri Wilensky All rights reserved See http ccl northwestern edu netlogo models Fire for terms of use The extension to the model was written by Bill Teahan To refer to this model in publications please use Firebreak NetLogo model Teahan W J 2010 Artificial Intelligence Ventus Publishing Aps Solution to Exercise 6 2 1 The reactive agent approach to this problem is for the agent to merely react to the situation it finds itself in A reactive agent does not require a representation of the universe in which it is operating It therefore cannot carry out the type of a priori reasoning required for the plan adopted by the cognitive agent as detailed in Exercise 6 2 1 One possible way to construct a reactive agent that can resolve this problem is to define a set of rules that the agent executes RI If am in front of the door and I have a key then open it R2 If am in fro
218. ules First Order Logic 1 Peter is an old man Fact database Peter is man Man Peter Old Peter Peter is old 2 Mary is the sister of Fact database Mary is sister Sister Mary Peter Peter of Peter 3 Some cats have large Fact database Unknown is 4x Cat x A Large _ears x ears cat Unknown has large ears 4 IfX is the sister of Y Rules database VxVy Sister x y A Boy Y gt and Y is a boy then Y JF x is sister of y and y is boy Brother y x bs the brother of X THEN y is brother of x 5 Simba is a white Fact database Simba is Elephant Simba White Simba elephant elephant Simba is white 6 All elephants are Rules database IF x _ is Vx Elephant x Mammal x mammals elephant THEN x is mammal 7 Elephants have a Rules database IF x is Yx Elephant x gt Has_trunk x A trunk and large ears elephant THEN x has trunk Has_large_ears x AND x has large ears 8 Ruby is a black and Fact database Ruby is cat Cat Ruby Black Ruby white cat Ruby is black Ruby is white White Ruby 9 Ruby is the mother of Fact database Ruby is mother Mother Ruby Pearl Pearl and Pearl is the of Pearl Pearl is daughter of Daughter Pearl Ruby daughter of Ruby Ruby 10 Some cats are small Fact database Unknown is 4x Cat x ASmall x cat Unknown is small 11 All cats eat meat Rules database IF x is cat Yx Cat x gt Eats_meat x Or THEN x eats meat Vx Cat x
219. uring the reasoning process Which are easy to follow and which are difficult Try finding out how the knowledge for the different types of knowledge representation is initialised Hint look at the setup knowledge command Find out how rules are defined and then how they are visualised as event maps Note that in the model rules and semantic networks are defined explicitly whereas frames and decision trees are setup by converting from rules How Have a go at explicitly defining the knowledge for frames and decision trees instead rather than use a conversion process EXTENDING THE MODEL Try adding other types of knowledge representation perhaps some that are hybrids of the ones implemented in the model For example the semantic event network type that comes with the model is a hybrid of the semantic networks and event map types Investigate alternative methods for visualising the reasoning process for each type of knowledge representations 6007 A a swasis yay 282 aus It s only an opportunity if you act on it IKEA SE STUDENT ven oes a Download free ebooks at bookboon com 141 Artificial Intelligence Exercises Agent Behaviour 9 Knowledge NETLOGO FEATURES Note the use of the hatch command to create walker agents to continue the walking of the event map network during the reasoning process How the walker agents walk the network determines the type of reasoning RELATED MODEL
220. utside complimented Al and pledged to work with President elect Bush Then Tony and Cherie accompanied Hillary Chelsea and me to the University of Warwick where I gave another of my farewell speeches this one on the approach to globalization our Third Way group had embraced trade plus a global contract for economic empowerment education health care and democratic governance The speech also gave me a chance to publicly thank Tony Blair for his friendship and our partnership I had treasured our times together and would miss them Download free ebooks at bookboon com 82 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication Before we left England we went to Buckingham Palace accepting Queen Elizabeth s kind invitation to tea We had a pleasant visit discussing the election and world affairs Then Her Majesty took the unusual step of accompanying us down to the ground floor of the palace and walking us out to our car to say good bye She too had been gracious and kind to me over the past eight years Describe how this excerpt reveals some of the thoughts of Bill Gates i e what he knows and doesn t know his beliefs his likes or dislikes his opinions his hopes and fears what he respects and has high regard for or what he disrespects and has contempt for what he supports or is opposed to his desires wants needs and his intentions Download free ebooks a
221. version A simple work around to this problem is to keep running the model using version 4 0 4 rather than 4 1 If you would like to load the Central Park Map shown in Chapter 9 of the Artificial Intelligence Book first ensure you are running the model in version 4 0 4 then select the Import option from the File tab select the Import World option then import the Map Drawing Central Park World file Map Drawing Central Park World csv Download free ebooks at bookboon com 125 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge Legend LAE yy setup Lt clear background draw path Sth St th St a import background fis m erase path m a draw object E E erase object via am st select object E select region E on fA erase region a a wm Ea St i igan m frat w w q ea 0 St Le Central Park Sout th St z 2 C u eee M Figure 9 7 1 2 Screenshot of the Interface for the Map Drawing model with the Central Park map loaded into the environment SimCorp is a leading provider of software solutions for the financial industry We work together to reach a common goal to help our clients succeed by providing a strong scalable IT platform that enables growth while mitigating risk and reducing cost At SimCorp we value commitment and enable you to make the most of your ambitions and potential Are you among the best qualified in finance
222. ween states States own three variables depth The depth in the event map tree stream The stream name where a stream consists of a sequence of sensory or motor events event The event either sensory or motor Abstract events are not needed for this particular event map A spring layout is used to visualise the event map HOW TO USE IT Press the setup button first This will usually produce a cluttered layout To unfold the clutter press the change layout button and then dynamically change the values in the sliders that control the layout One effective technique is to reduce the value of the spring length slider to 0 then slowly increase it back up again until the desired length and layout is achieved THE INTERFACE The model s Interface buttons are defined as follows setup This will clear the environment and variables and re load the event map Normally this will appear in a cluttered form and the change layout button needs to be pressed subsequently change layout This can be used to clear some of the clutter by changing the values in the three sliders The model s Interface sliders are defined as follows spring constant This is a value used by the layout spring command Changing it will usually not affect the visualisation of the event map much Download free ebooks at bookboon com 134 Please click the advert Artificial Intelligence Exercises Agent Behaviour 9 Knowledge
223. what you want out of your future per an employer with a broad range of operations in a load of countries will always pe the ticket Working within the Volvo Group means more than 100 000 friends and B tieagues in morethan 185 countries all over the world We offer graduates great career opportunities check out the Career section at our web site www volvogroup com e look forward to getting to know you VOLVO AB Volvo publ Were woNOQrOUR COm Vowo Teucxs Rexauer Trucns Macx Trucxs Vowo Buses Vowo Coxsteucrion Equipwext Vowo Pena Vowo Aero Vowo IT Vowo Financat Sernces Vowo 3P Vowo Powenreain VowoPaers Vowo Tecumoocr VowoLoasnes Business Anca Asa Download free ebooks at bookboon com 195 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Action Add X to database Cycle 3 Database ABCODVY X Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Action Add Z to database Cycle 4 Database ABCVYX 2 Goal Z reached Answer for ii Found rules with goal or sub goal marked by Proven rules marked by Step 1 Goal Z Database ABCD Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3 V amp C gt X Rule 4 D gt V Step 2 Sub Goal X from Rule 1 Database ABCD Rules base Rule 1 X amp Y amp ADZ Rule 2 V amp B gt Y Rule 3
224. while at the same time spreading the fire from the previously selected trees When the satellite image is imported this will end up showing which parts of the image the is t ree reporter considers to be trees ignite forest This will set an ignition point where the mouse is clicked from which the fire will start burning Keeping the mouse down will draw multiple overlapping ignition points The fire will then spread out once the go button is pressed Download free ebooks at bookboon com 160 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises ignite on the left The vertical line of fire will be ignited from the left side of the environment The fire will then spread across to the right once the go button is pressed draw firebreak This will draw a firebreak coloured dark brown into the environment at the point where the mouse is clicked To continue drawing the firebreak keep on clicking the mouse while at the same time dragging it in the desired direction remove firebreaks This will remove all firebreaks that have been drawn in the environment The Interface monitor and sliders are defined as follows percent burned This is the percentage of trees that have been burnt density This is the density of the randomly generated forest when the setup button is pressed firebreak width This sets the width of the firebreaks fire fade amount This controls how long a tree
225. wing six stages tokenisation gt extraction of complex words gt processing of basic phrases gt processing of complex phrases extraction of semantic patterns gt merging of semantic patterns into templates For question answering there is the added step of applying the information extraction to the question text and matching the extracted templates However there are still major hurdles that need to be overcome if information extraction and question answering technology are to be applied successfully to the problem of designing a conversational agent Some of the hurdles that need to be overcome before the agent is believable in the sense that it can pass the Turing Test are as follows e Jt must understand and be able to employ common sense reasoning e It must have access to world knowledge e It needs to have some form of understanding of the real world Brooks argues that an A I will have to be physically embodied in the real world e It needs to understand the context of the conversation and respond accordingly e It needs to be able to answer questions such as What are you thinking about in a realistic way e And so on Perhaps the main shortcoming of present day conversational agents is the lack of understanding of the context of the conversation what has been said before what needs to be said next what is appropriate and so on Part of the problem is the need for access to general knowledge as well as common s
226. wledge using maps 124 9 8 Representing knowledge using event maps 133 9 9 Representing knowledge using rules and logic 143 9 10 Reasoning using rules and logic 144 9 11 Knowledge and reasoning using frames 145 9 12 Knowledge and reasoning using decision trees 145 9 13 Knowledge and reasoning using semantic networks 146 10 Intelligence 147 10 1 The nature of intelligence 147 10 2 Intelligence without representation and reason 147 A You re full of energy s 3 a and ideas And that s just what we are looking for Looking for a career where your ideas could really make a difference UBS s Graduate Programme and internships are a chance for you to experience for yourself what it s like to be part of a global team that rewards your input and believes in succeeding together Wherever you are in your academic career make your future a part of ours by visiting www ubs com graduates www ubs com graduates a UB S Download free ebooks at bookboon com Please click the advert Artificial Intelligence Exercises Agent Behaviour Contents 10 3 What AI can and can t do 148 10 4 The Need for Design Objectives for Artificial Intelligence 148 10 5 What are Good Objectives 148 10 6 Some Design Objectives for Artificial Intelligence 148 10 7 Towards believable agents 149 10 8 Towards computers with problem solving ability 152 Solutions to Selected Exercises 158 goo n a swasis yay 20u SN 3 It s only an opportu
227. work with a cone of vision and particle agents that simulate the velocity of each particle and takes into consideration viscosity wind and gravity None of these solutions work i e they behave nothing like water does HOW TO USE IT Press the setup environment button in the Interface first to reset the simulation Then choose one of the solutions using the solution chooser See what happens when you press the go simulation button THE INTERFACE The model s Interface buttons are defined as follows setup simulation This will reset the simulation back to the start state which is where all the reservoirs have no water in them and the tap is turned off go simulation This turns on the tap and agents that simulate water droplets start flowing downwards out of the tap into the first reservoir on the left The model s Interface sliders switches and chooser are defined as follows 1 For when the solution variable is set to either turtle agents 1 or turtle agents 2 drops per tick This is the number of drops that will emerge out of the tap every tick gize of drop This is the size of each drop shape of drop This is the shape of each drop Possible shapes are drop circle square and hex 2 For when the solution variable is set to boid agents 1 vrate of random turn This controls how much the boid randomly turns each tick radius angle and radius length This defines t
228. xpand your thinking and provide you with the foundations for a successful career in business The programme is developed in consultation with recruiters to provide you with the key skills that top employers demand Through 11 months of full time study you will gain the business knowledge and capabilities to increase your career choices and stand out from the crowd London Business School Regent s Park London NW1 4SA Applications are now open for entry in September 2011 United Kingdom Du P Tel 44 0 20 7000 7573 For more information visit www london edu mim Email mim london edu email mim london edu or call 44 0 20 7000 7573 www london edu mim Download free ebooks at bookboon com 201 Artificial Intelligence Exercises Agent Behaviour Soulutions to Selected Exercises Solution to Exercise 10 7 2 There are many shortcomings of Eliza like solutions such as that provided in the Chatbot NetLogo model which is based on using regular expressions to find basic patterns in the text However there are more sophisticated approaches that are based on regular expressions that hold promise A common solution used for information extraction for example is that each stage of linguistic processing is based on regular expressions that extract some information from the text This is then passed on to the next stage eventually ending up with a set of templates as the output The Fastus system for example uses the follo
229. y Biology gt Heatbugs http ccl northwestern edu netlogo models Heatbugs Initial settings for bugs Wd ticks 232 3D bug count 100 bugs min ideal temp 10 min output heat 5 max ideal temp 40 max output heat 25 Actions setup go g deep freeze heat up Other parameters OK to change during evaporation rate 0 01 run aS I diffusion rate 0 9 i random move chance 0 Avg Bug Unhappiness N Se w ideal temp happiness Color bug by 0 time 251 watch saddest watch happiest reset perspective Figure 6 6 1 Screenshot of the Interface for the Heatbugs model after the setup button has been pressed followed by the go button with the slider values as shown in the image o unhappiness Download free ebooks at bookboon com 19 Please click the advert Artificial Intelligence Exercises Agent Behaviour 6 Behaviour The model s Information states that it has been used as a demonstration model for many agent based modeling toolkits with several different kinds of emergent behaviour resulting from simple rules Run the model multiple times to observe what happens for the different slider values Make a list of the emergent behaviours you observe How is the temperature diffused throughout the environment in this model Each tick all the turtles are directed to perform the step procedure Where is this done in the model Explain what happens when the step
230. y and code length is the one that usually gets close to the source entropy and code length whereas the other agents entropy and code lengths reflect how different the slider settings are from the source distribution WHAT IS ITS PURPOSE The purpose of this model is show how entropy and code length calculations are made given a probability distribution Download free ebooks at bookboon com 67 Please click the advert Artificial Intelligence Exercises Agent Behaviour 7 Communication HOW IT WORKS A car turtle agent is used to represent the cars in the environment These are created with colours distributed according to the source distribution The arrival of the next car to the left of the environment is determined by a random number generated according to the slider next car random tick interval Turtle agents called agents are also present but are not shown in the animation These can be considered to be observers watching the cars go past Each owns a distribution which is a list of counts that are used to calculate the probabilities in order to guess the upcoming cars HOW TO USE IT Setting the values on the sliders for the Source Distribution will determine the distribution for the cars that appear in the animation Setting the next car random tick interval will control how often the cars appear The user can then set the slider values for the three Agent distributions on the left to see how this affec

Artificial Intelligence - Kenyatta University Library

Contents

Download Pdf Manuals

Related Search

Related Contents