Evolving Self Organizing Maps User Manual
Contents
1. 1 QE Vi allWemui Kill 7 where WBMULW weights vector of BMU i N number of pattern of dataset x input vector i assigned to current BMU The equation 7 corresponds to the average of distance of each pattern form its BMU ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved E waRE LI DAta Mining amp Exploration po Program Pa a Weng Pt cedettero 2 3 Clustering quality indicator The results of clustering process can be evaluated using the Davies Bouldin DB index This index measures the ratio of intra cluster and extra cluster distances measured from centroids Davies amp Bouldin 1979 The internal scatter of a cluster C can be written as Si Yee le 211 4 i 1 K 8 Lq ICil X ECi l where C is the number of pattern assigned to cluster i x and z are respectively a pattern of cluster i and his centroid g is an absolute value K is total number of clusters The distance between two clusters can be written as o 1 t dije Z z Pila zs 9 where z e z represent respectively centroids of clusters i and jJ Zsi Z denotes the absolute value of the difference between vectors z and z computed on dimension s D is the total number of pattern t is an absolute value So the DB index can be written as DB EI maxy jei fetta 10 Low values of this index indicate a better clustering Howe2. features BMU cluster and activation of winner node E_SOM_ Train_Normalized_Results txt File with same structure of The file is produced only if E_SOM_ Test_Normalized_Results txt precedent described file but normalization of dataset was E_SOM_Run_Normalized_Results txt with normalized features requested E_SOM_ Train_Histogram png E_SOM_Test_Histogram png Histogram of clusters found E_SOM_Run_Histogram png E_SOM_Train_Validity_Indices txt E_SOM_Test_Validity_Indices txt E_SOM_Run_Validity_Indices txt Quantization error and DB index are always produced ICA and ICC are produced File that reports the validity indices of the experiment ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved Pass pre i fy vs phe dw Sees ws DAta Mining amp Exploration Program only in Test use case E_SOM_Train_U_matrix png E_SOM_Test_U_matrix png U Matrix image E_SOM_Run_U_matrix png nemere The Uheight value is used to E_SOM_Train_Output_Layer txt output layer reports ID Sa ssa E_SOM_ Test_Output_Layer txt coordinates clusters number E_SOM_Run_Output_Layer txt of pattern assigned and Uheight value File that for each clusters E_SOM_Train_Clusters txt reports label DIED EE Ok E_SOM_Test_Clusters txt peo assigned PARA E SOM Run Clusters txt of association respect total ca number of pattern and its centroids F 5OM Tram Clusiered mage p ng I
3. sccsissctarasasacacaiaacdasccensiavedensdaeaianeceasiesdecnsconsseualaaddneaiioneaasiendenessansess 14 TABLE INDEX TOUT OU A NI 10 Table 2 List of model parameter setup web help pages available iii 10 TGDIC 3 ADD TEVIGUONS and GOT ON iii 14 TD iC CHC OCI Riano 15 Tabled Applicable DOCUMECIIS an civcnnvnvadsedsassavnacevaiseowiarnorndadivuieusaveenesestesRseiseouleyabiavines OESS OAE 16 FIGURE INDEX Figure 1 Flow chart of a generic Unsupervised neural network iii 4 Figure 2 Connected nodes reveals ClUSters iii 6 RESOR E 6 Figure 4 Modified U Matrix for ESOM model 7 Figure 5 The starting point with a Workspace esomExp created and input dataset uploaded Il Figure 6 Selection of functionality and USC Case iii Il Figure 7 The EsomIris experiment configuration tab iii 12 Figure 8 Experiment finished MeSSAC ccccccccccccccnveccceccceeeeneeeeseeeeeeeee a eeeeeeeeeeeaaaaeeeeeeeeeeeaaaaaeseeeeesesaaaaaseeees 12 Fiore 9 Listo ouput ile Produ Ed asserit snene En rE EEEE NEEE E 12 Figure 10 Moving configuration file in the Workspace and uploading of target clusters file 13 2 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 1 Introduction he present document is the user guide of
4. the data mining model Evolving Self Organizing Maps ESOM a data mining model that can be used to execute scientific experiments for clustering on massive data sets formatted in one of the supported types ASCII columns separated by spaces CSV comma separated values FITS Table numerical columns embedded into the fits file VOTable GIF JPG and FITS Image This manual is one of the specific guides one for each data mining model available in the webapp having the main scope to help user to understand theoretical aspects of the model to make decisions about its practical use in problem solving cases and to use it to perform experiments through the webapp by also being able to select the right functionality associated to the model based upon the specific problem and related data to be explored to select the use cases to configure internal parameters to launch experiments and to evaluate results The documentation package consists also of a general reference manual on the webapp useful also to understand what we intend for association between functionality and data mining model and a GUI user guide providing detailed description on how to use all GUI features and options So far we strongly suggest to read these two manuals and to take a little bit of practical experience with the webapp interface before to explore specific model features by reading this and the other model guides All the cited documentation package is available f
5. DAta Mining amp Exploration Program Lit MA parti Scienze 2 ee DI mento di Fisiche OF ISTITUTO NAZIONALE di ASTROFISICA f CI 6 CALTECH Universita di Napoli Federico H Pre OSSERVATORIO ASTRONOMICO di CAPODIMONTE el a sp Te 7 be k By ye a Ve gt E gt Sul so at One e y __ c i D asm b p 4 LP a re gt i d _ lt a om a _ ee i c s gt A n g J A i x x LI f e 3 x 7 E 4 et at y lt s KAKAK DI Evolving Self Organizing Maps User Manual DAME MAN NA 0021 Issue 1 2 Author M Brescia F Esposito Doc ESOM_UserManual_ DAME MAN NA 0021 Rel1 2 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Index L POG CON i ilo 3 2 PSOM CIC ORCC A OY CM ised rie avo 4 2A MMS AICS RESO RE RR RR TRO 5 2 1 1 ESOM output layer visualization 6 0888 EER RA 7 2 2 1 Onana on E 0 7 23 lostenn GUA IMI Calorie re lidia a 8 eI Tore site Jk Item neuen pene ne RAD Tt E eo nT an nn E R T 8 LL Ipo tase E S sac tana coew EAA E EAE AE cocscaentenee 9 005 00 PR 0 9 39 Expe ment parameler SS PRORPIO RIA 10 Be AIC RR E E O 11 ac Piisterample Mis Wat AS scire rien 11 dall U C aa 1 12 Estela eo 12 5 Appendix References and ACrOny Mis
6. _Configuration txt in the Workspace Moreover in order to execute a Test we need a file with one single column with the target clusters of each pattern Also this file must be uploaded in the Workspace 12 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Norkspace somExp i Dow Edit File w Type LastAccess Ci amp E_SOM_Train_Network_Configuration txt other 2013 09 04 3 iris ascii 2013 09 04 we _ iris_target txt ascii 2013 09 04 Figure 10 Moving configuration file in the Workspace and uploading of target clusters file Now we have to create a new experiment and choose the functionality Clustering _E_ SOM and select Test as use case For this model test has only five mandatory parameters e input file iris txt e configuration file file produced by a Train use case which contains experiment parameters e dataset target file file that report the cluster of each pattern present in the input dataset e dataset type 0 which indicates and ASCII input file Workspace esomExp Select a Running Test Experiment esomTest Mode M ento Clustering_E_SOM m _ en Field is Required input file iris ba x configuration file E_SOM_Train_Network_ w dataset target file iris_target ixt v Submit After submission the experiment will be executed and will produced the output file exp
7. ected 13 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved pm WARE 4 vs Phe de Sees wwe _ bal y SEC DAta Mining amp Exploration Program 5 Appendix References and Acronyms Abbreviations amp Acronyms A amp A Al ANN ARFF ASCII Bok BP BLL CC CSOM CSV DAL DAME DAMEW ARE DAPL DL DM DMM DMS FITS FL FW GRID GSOM GUI HW Meaning Artificial Intelligence Artificial Neural Network Attribute Relation File Format American Standard Code for Information Interchange Base of Knowledge Back Propagation Business Logic Layer Connected Components Clustering SOM Comma Separated Values Data Access Layer DAta Mining amp Exploration DAME Web Application REsource Data Access amp Process Layer Data Layer Data Mining Data Mining Model Data Mining Suite Flexible Image Transport System Frontend Layer FrameW ork Global Resource Information Database Gated SOM Graphical User Interface Hardware A amp A KDD IEEE INAF JPEG LAR MDS MLC MLP MSE NN OAC PC PI REDB RIA SDSS SL SOFM SOM SW TWL UI URI VO XML Meaning Knowledge Discovery in Databases Institute of Electrical and Electronic Engineers Istituto Nazionale di Astrofisica Joint Photographic Experts Group Layered Application Architecture Massive Data Sets Multi Layer Clustering Multi Layer Perceptron Mean S
8. he Growing Cell Structures GCS that introduced the incremental aspect of the network preserving a connection between nodes One year later Fritzke 1995 the Growing Neural Gas GNG remove also this aspect In the same year Bruske amp Somemr 1995 introduced the Dynamic Cell Structures GCS DCS GCS which differing from GNG slightly in the location of node insertion However all the models described each one of them for different reasons implied additional computational time which can be reduced as proposed in the Evolving Self Organizing Maps Deng amp Kasabov 2003 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved WARE EE DAta Mining amp Exploration _ Program 2 1 The model ESOM The algorithm starts with a null network without any nodes Nodes are created incrementally when a new input pattern is presented the prototype nodes in the network compete with each other and the connections of winner node are updated In particular if the two winners are not connected a connection will be made between them New node will be inserted into the network if none of existing nodes matches with the current input In this case the new node also sets connections to the first two winners Let it be x input pattern W w wWa Wy existing prototype set E minimum distance threshold A new node is inserted if w gt Cj Vw EW 1 and it is initia
9. he weakest connection is pruned and this process goes on during the whole dataset ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Cluster B Cluster A Figure 2 Connected nodes reveals clusters 2 1 1 ESOM output layer visualization The standard tool for visualization and interpretation of a SOM is the U Matrix For each node of Kohonen layer a value will be computed according to its distance to adjacent nodes This value can be visualized on a heat map in which light colours represents nearby nodes in the weights space while dark colours represents distant nodes Moutarde amp Ultsch 2005 Typically the map is represented on a greyscale as shown in Figure 3 In order to increase further the interpretability of U Matrix is possible to overlay to each node BMU of some pattern a colour that identify the relative cluster RI LI e aP O0 _ EL EE 0O nm ee Figure 3 Example of U Matrix Since in ESOM model neurons are not placed on a rigid structure such as the grid of Kohonen layer it s evident the impossibility to use the classical U Matrix as visualization tool However in order to provide a method to visualize the clustering results a modified U Matrix has been implemented In this type of U Matrix neurons are not arranged on the grid depending on the actual position in parameter space but gr
10. lized as Wy 1 X 2 The eq 2 show that a new node is inserted representing exactly the poorly matched input vector This approach leads to a computational efficiency because other type of insertions as the mid point insertion used in GNG takes a greater number of iterations Although direct allocation in ESOM is sensitive to noise and may introduce some artefacts in clustering this can be mitigated by automatic deletion of obsolete nodes When an input pattern matches well with some prototype the activation of the winner node is defined as 2 2 x w ae 3 In the ESOM model the neighborhood of a node is defined as Q U s i j gt 0 4 where s i j represent the weight of connection between nodes i and j and the neighborhood function can be written as _ a x Nip X Lik A X g The weights update follow the formula sm f t x wi ifi 005 YK ak 6 0 else where y learning rate typically constant set to 0 05 b BMU In eq 6 clearly shows the strong analogy with classic Kohonen learning rule in which change only the definition of neighborhood The neighborhood concept used in ESOM based on connection result computationally less expensive than other methods as the rank used in GNG and allows to visualization of clustered structure of data Figure 2 In order to do this a mechanism to delete the weak connection is required After the presentation of an established number of pattern t
11. m DAME Program we make science discovery happen REL 18 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved
12. mage that show the effect of The file is produced only 1f POM Tesis lugterod Image png the clustering process input dataset is an image E_SOM_Run_ Clustered_image png E_SOM_ Train_Clustered_image txt File that for each pixel a E_SOM_ Test_Clustered_image txt reports ID coordinates Do Bis p cue oniyi E_SOM_Run_Clustered_image txt features and cluster assigned MIRO ataselns aly ABs E_SOM_Train_Datacube_image zip Archive that includes the The file is produced only if E_SOM_Test_Datacube_image zip clustered images of each i i E_SOM_Run_Datacube_image zip slice of a datacube i Calan Table 1 Output file list coordinates are computed according to cluster membership 3 3 Experiment parameter setup There are several parameters to be set to achieve training specific for network topology and learning algorithm setup In the experiment configuration there is also the Help button redirecting to a web page dedicated to support the user with deep information about all parameters and their default values We remark that all parameters labeled by an asterisk are considered as required In all other cases the fields can be left empty default values are used and shown in the help web pages The following table reports the web page addresses for all clustering models and related use cases subject of this manual Functionality Model USE SETUP HELP PAGE CASE http dame dsf unina it clustering esom html A http dame dsf unina it clu
13. networks for Kiang M Y 2001 clustering analysis Computational Statistics amp Data Analysis Vol 38 161 180 Topology preservation in self organizing maps Kiviluoto K 1996 Proceedings of the International Conference on Neural Networks 294 299 Self Organizing Maps 3 ed Springer Kohonen T 2001 U F Clustering A new performant cluster mining method on Moutarde F Ultsch A 2005 segmentation of self organizing map Proceedings of WSOM 05 September 5 8 Paris France 25 32 Clustering with SOM U C Proc Workshop on Self Ultsch A 2005 Organizing Maps Paris France 75 82 Clustering of the Self Organizing Map IEEE Transactions Vesanto J Alhoniemi E 2000 on neural networks Vol 11 No 3 586 600 A K means clustering algorithm Applied Statistics 28 Hartigan J A Wong M A 1979 100 108 Table 4 Reference Documents ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved ID AI A2 A3 A4 AS A6 A7 AS A9 A10 AII AI2 A13 Al4 AI5 A16 A17 A18 A19 Program Title Code Author SuiteDesign_VONEURAL PDD NA 0001 Rel2 0 DAME Working Group project_plan_ VONEURAL PLA NA 0001 Rel2 0 Brescia statement_of_work_VONEURAL SOW NA 0001 Rell 0 Brescia mlpGP_DAME MAN NA 0008 Rel2 0 Brescia pipeline_test_VONEURAL PRO NA 0001 Rel 1 0 D Abrusco scientific_example VONEURAL PRO NA 0002 Rel 1 1 D Abrusco Cavuoti fr
14. of networks with supervised training The main use of these networks is precisely the data analysis in order to found groups having similarities pre processing and data clustering or form classification recognition of images or signals The supervised learning consists in the training of a network by input target pairs that obviously are knows solutions of optimization problems in specific points of data space parameters space of problem itself classification approximation or functions regression Sometimes there is not the possibility to have data relative to solution of problems but data to analyse without specific information on them unsupervised training A typical problem of such type is the research of class or groups of data with similar features within an unordered group of data clustering Generally clustering problems needs the use of a competitive rule among the nodes of the network in which the winner is candidate to represents the input pattern In the most well known self organizing neural network SOM Kohonen 2001 the nodes are placed on the top of a grid forming a two or three dimensional topologically constrained space The principal limitation of this type of structure is the static of the output layer that especially results a problem in case of on line clustering in which is useful a network capable of evolve itself as new data are acquired During the years different solution were proposed Fritzke 1994 propose t
15. ontend_VONEURAL SDD NA 0004 Rel1 4 Manna FW_VONEURAL SDD NA 0005 Rel2 0 Fiore REDB_VONEURAL SDD NA 0006 Rel1 5 Nocella driver VONEURAL SDD NA 0007 Rel0 6 d Angelo dm model_ VONEURAL SDD NA 0008 Rel2 0 Cavuoti Di Guido ConfusionMatrixLib_VONEURAL SPE NA 0001 Rel1 0 Cavuoti softmax_entropy_VONEURAL SPE NA 0004 Rel1 0 Skordovski Clustering con Modelli Software Dinamici Seminario Dip Esposito F di Informatica Universita degli Studi di Napoli Federico II http dame dsf unina it documents html dm_ model VONEURAL SRS NA 0005 Rel0 4 Cavuoti DMPlugins_DAME TRE NA 0016 Rel0 3 Di Guido Brescia BetaRelease_ReferenceGuide DAME MAN NA 0009 Brescia Rell 0 BetaRelease_GUI_UserManual DAME MAN NA 0010 Brescia Rel1 0 SOM and 2 stage clustering models Design and Esposito Brescia Requirements som _DAME SPE NA 0014 Rel4 0 Table 5 Applicable Documents ESOM Model User Manual DAta Mining amp Exploration Date 15 10 2008 19 02 2008 30 05 2007 04 04 2011 17 07 2007 06 10 2007 18 03 2009 14 04 2010 29 03 2010 03 06 2009 22 03 2010 07 07 2007 02 10 2007 2013 05 01 2009 14 04 2010 28 10 2010 03 12 2010 2013 This document contains proprietary information of DAME project Board All Rights Reserved 16 DAta Mining amp Exploration Program 000 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Progra
16. ouped by cluster membership ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program ili Lilli 2 ost 0 1825 ag ia eel E a 0 185 16 Figure 4 Modified U Matrix for ESOM model As can be seen in Figure 4 the number of identified clusters is immediately clear as well as the number of nodes assigned to each of them As one of the peculiarity of this model is the presence of weighted connections between the output nodes we have choose to show the average of weights of connections of each node as gradient on gray scale So dark nodes are node without connections or with very weak ones 2 2 SOM quality indicators Good criteria to evaluate the quality of aSOM were proposed by Kiviluoto 1996 1 What is the degree of continuity for the map topology ii What is the resolution of the map topology A quantification of these two properties can be obtained by computation of quantization error and topographic error Chi amp Yang 2008 However the lack of classic grid of the output layer does not allow to evaluate the topographic error as it has been defined Thus the ESOM model provides only the quantization error as a quality criterion 2 2 1 Quantization error The quantization error is used to the computation of similarity of pattern assigned to the same BMU according to the following formula
17. quare Error Neural Network Osservatorio Astronomico di Capodimonte Personal Computer Principal Investigator Registry amp Database Rich Internet Application Sloan Digital Sky Survey Service Layer Self Organizing Feature Map Self Organizing Map Software Two Winners Linkage User Interface Uniform Resource Indicator Virtual Observatory eXtensible Markup Language Table 3 Abbreviations and acronyms ESOM Model User Manual 14 This document contains proprietary information of DAME project Board All Rights Reserved F wane 4 i i A DAta Mining amp Exploration Program Veg A ke Sees ee Reference amp Applicable Documents Title Code Author Date Dynamic cell structure learns perfectly topology preserving Bruske J Sommer G 1995 map Neural Comput 7 845 865 A Two stage Clustering Method Combining Ant Colony SOM Chi S C Yang C C 2008 and K means Journal of Information Science and Engineering 24 1445 1460 A cluster separation measure IEEE Transactions on Davies D L Bouldin D W 1979 Pattern Analysis and Machine Intelligence Vol 1 224 227 On line pattern analysis by evolving self organizing maps Deng D Kabasov N 2003 Neurocomputing 51 Elsevier 87 103 Improved interpretability of the unified distance matrix with Hamel L Brown C W 2011 connected components Proceedings of the 2011 International Conference on Data Mining Extending the Kohonen self organizing map
18. rom the address http dame dsf unina it dameware html where there is also the direct gateway to the webapp As general suggestion the only effort required to the end user is to have a bit of faith in Artificial Intelligence and a little amount of patience to learn basic principles of its models and strategies By merging for fun two famous commercial taglines we say Think different Just do it casually this is an example of data text mining ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program 2 ESOM theoretical overview The goal of this guide is to show the use of the unsupervised model for clustering ESOM START Submit a training Reset back to first pattern to the training pattern to try network again Any more e W Increase epoch ee counter by 1 l Adjust weights based on learning rule STOP Figure 1 Flow chart of a generic unsupervised neural network The theory of neural network is based on computational models introduced in 40s by McCulloch amp Pitts 1943 which reproduced in a simplified way the behaviour of a biological neuron The neural networks are self adaptive computational models based on the concept of learning from examples supervised or self organizing unsupervised The self organizing neural networks are suitable for the solution of different problems in respect
19. stering esom html train CAUS TE ees http dame dsf unina it clustering_esom html test http dame dsf unina it clustering_esom html run Table 2 List of model parameter setup web help pages available 10 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved WARE amp DAta Mining amp Exploration A Program Vey Pt 60 Sees eer 4 Examples This section is dedicated to show some practical examples of the correct use of the web application Not all aspects and available options are reported but a significant sample of features useful for beginners of DAME suite and with a poor experience about data mining methodologies with machine learning algorithms In order to do so very simple and trivial problems will be described Further complex examples will be integrated here in the next releases of the documentation 4 1 First example Iris Dataset This example shows the use of the ESOM model applied to the dataset Iris The first step consists in the creation of a new workspace named for example esomExp and the input dataset iris txt must be uploaded in the workspace just created Workspace v File Manager Workspace oa New Workspace is Plot Editor f Image Viewer esomExp C Dow Edit File Type Last Access Rename Workspace C Upload jf Experiment 3 Delete eet o iris txt ascii 2013 09 04 P somExp amp ia x esomExp i dii x Figure 5 The star
20. ting point with a Workspace esomExp created and input dataset uploaded 4 1 1 Train Use Case Let suppose we create an experiment named EsomIris and we want to configure it After creation the new configuration tab is open Here we select Clustering_ E_SOM which indicates the functionality and the model We select also Train as use case Workspace esomExp Select a Running Frain x Experiment Esomiris Mode S i select a Clustering E_SOM Functionality Field is Required Figure 6 Selection of functionality and use case Now we have to configure parameters for the experiment In particular we will leave empty the not required fields labels without asterisk As alternative you can click on the Help button to obtain detailed parameter description and their default values directly from the web application We give iris txt as training dataset specifying e dataset type 0 which is the value indicating an ASCII file e input nodes 4 because 4 are the columns in input dataset e epsilon 0 5 e pruning frequency 10 11 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved DAta Mining amp Exploration Program Note that the values of epsilon and pruning frequency can have a great influence on the results of the experiment Unfortunately these values can be only set by a try amp error process Workspace esomExp Select a Running Trai i rain
21. twork with new datasets The Run use case implies the simple execution of the trained and tested model like a generic static function 3 1 Input We also remark that massive datasets to be used in the various use cases are and sometimes must be different in terms of internal file content representation Remind that it is possible to use one of the following data types e ASCII extension dat or txt simple text file containing rows patterns and columns features separated by spaces e CSV extension csv Comma Separated Values files where columns are separated by commas e FITS extension fits or fit fits files containing images and or tables VOTABLE extension votable formatted files containing special fields separated by keywords coming from XML language with more special keywords defined by VO data standards e JPEG extension jpg or jpeg image files e PNG extension png image files GIF extension gif image files 3 2 Output In terms of output the following file are obtained FILE DESCRIPTION REMARKS File containing the Must be moved to File E_SOM_Train_Network_Configuration txt parameters of a trained Manager tab to be used for network test and run use cases E_SOM_Train_Status log E_SOM_Test_Status log E_SOM_Run_Status log E_SOM_Train_Results txt E_SOM_Test_Results txt E_SOM_ Run_Results txt File containing details on the executed experiment File that for each pattern reports ID
22. v Experiment Esomiris Mode Bison Clustering_E_SOM OS unclionalty Field is Required input file iris txt v configuration file v dataset type 0 input nodes 4 normalize data learning rate epsilon 0 5 pruning frequency 10 Submit Figure 7 The EsomIris experiment configuration tab After submission the experiment will be executed and a message will be shown when the execution is completed Workspace v File Manager 1 4 Workspace New Workspace Uda Plot Editor Image Viewer isomEXp E Dow Edit File Type Last Access Rename F Workspace E Upload f Experiment Delete i Di ristxt ascii 2013 09 03 f TestSOM B di x Note Xx f TestESOM C e Experiment Finished j Please referto somExp workspace for results f somExp C OK Figure 8 Experiment finished message The list of output files obtained at the end of the experiment available when the status is ended is shown in the dedicated section Each file can be downloaded or moved in the Workspace Figure 9 List of output file produced 4 1 2 Test Use Case In this paragraph is shown how execute a Test Use Case starting from a Train previously executed Test use case is useful to evaluating the executed clustering by the indices described in paragraph Errore L origine riferimento non stata trovata In order to do this referring to the example shown above we have to move the file E SOM_Network
23. ver note that on non linearly divisible dataset could not be objective A more objective evaluation can be obtained if the cluster of each input data is known In such case is possible to computes the Index of Clustering Accuracy ICA and the Index of Clustering Completeness ICC Let it be NC number of tehoretical clusters NC number of clusters found NCa number of disjoint clusters Two theoretical clusters are disjoint if the intersection of the label assigned by clustering process in the two clusters is the empty set ICA INCc NCel 11 NC NCy Icc 1 4 11 NC Low values of these indices reflects best results 3 Use of the ESOM For the user the ESOM offer three use cases e Train e Test e Run 8 ESOM Model User Manual This document contains proprietary information of DAME project Board All Rights Reserved Pass pre i fy DAta Mining amp Exploration Program i n Sa Weng Pt cedettero Additionally to use cases just described is possible to perform a Train starting form a previously trained network This use case is called Resume Training A typical complete experiment consists of the following steps 1 Train the network with a dataset as input then store as output the final weight matrix best configuration of trained network weights 2 Test the trained network with a dataset containing both input and target features in order to verify training quality 3 Run the trained and tested ne
Download Pdf Manuals
Related Search