the PDF - Support
Contents
1. Model Evaluator stage within a dataflow that is configured to use the model 1 14 Using the View Contents of Advanced Analytics Repository on page 12 procedure navigate to the Repository page The Repository page displays models stored within the Advanced Analytics Repository via a model table If a large number of models are stored within the Advanced Analytics Repository then the model table may display models across several pages Using the model table select models to be deleted from the Advanced Analytics Repository Note Only models selected on the model table s current page will be deleted For example if the model table has two pages of models and models have been selected on both pages then clicking the delete button will only delete the models on the current page and ignore any models selected on the other page Click the Delete model button located directly above the model table The Delete model button will be disabled when no models are selected on the model table s current page A pop up message is displayed asking for confirmation to delete the selected model or models from the Advanced Analytics Repository Click the Yes button to confirm and delete the selected models from the Advanced Analytics Repository Click the No button to cancel the delete operation After successfully deleting the selected models from the Advanced Analytics Repository the Repository page will refresh and update the model table Spectru2. Spectrum Technology Platform Version v9 0 SP3 Analytics Guide Contents Chapter T IntroduclON iia cc 5 Advanced Analytics 6 Advanced Analytics Module Components uusnsresnnnennnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnn nn 6 Chapter 2 Model Evaluator 7 00 0 8 Deploying a 1000 44420554 ici ii 8 Re configure Model Evaluator Settings uunennnnnnnennnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnnannnnnn 9 Chapter 3 Advanced Analytics Repository ocoooncnonnnccnccccnnncnnananncnnnnnnannnnans 11 Introduction to the Advanced Analytics 0510 2 36 12 View Contents of Advanced Analytics 2051 0 36 12 View AAA EnheHn nn aha ne anne aaeb piia 12 Add a Model to the Advanced Analytics 3620540 13 Delete a Model from the Advanced Analytics 362050 14 Chapter 4 Supported Model 15 Introduction to Supported Model Formats 2 222202020002000nnnnnnunn ann 16 0 16 16 P MMs rior A AA A aia 17 Introduction In this section e Advanced Analytics Module ooooooo 6 e Advanced Analytics Module Components 6 Advanced Analytics Module Advanced Analytics Module The Spectrum Technology Platform Advanced Analytics Module evaluates statistical and predictive models defined in either QMML Portrait Miner s proprietary model format or PMML industry standard Predictive Model Markup Language It can be used to enri
3. VTEQ All rights reserved Data 2014 NAVTEQ North America LLC Fuente INEGI Instituto Nacional de Estadistica y Geografia Based upon electronic data National Land Survey Sweden Copyright United States Census Bureau Copyright Nova Marketing Group Inc Portions of this program are Copyright 1993 2007 by Nova Marketing Group Inc All Rights Reserved Copyright Second Decimal LLC Copyright Canada Post Corporation This CD ROM contains data from a compilation in which Canada Post Corporation is the copyright owner 2007 Claritas Inc The Geocode Address World data set contains data licensed from the GeoNames Project www geonames org provided under the Creative Commons Attribution License Attribution License located at http creativecommons org licenses by 3 0 legalcode Your use of the GeoNames data described in the Spectrum Technology Platform User Manual is governed by the terms of the Attribution License and any conflict between your agreement with Pitney Bowes Software Inc and the Attribution License will be resolved in favor of the Attribution License solely as it relates to your use of the GeoNames data ICU Notices Copyright 1995 2011 International Business Machines Corporation and others All rights reserved Permission is hereby granted free of charge to any person obtaining a copy of this software and associated documentation files the Software to deal in the Software without restric
4. ation For example rename a Spectrum Field Name or change its Include checkbox status to be included excluded from the dataflow Analytics Guide 9 Re configure Model Evaluator Settings 10 Any changes to the outputs must adhere to the following validation rules At least one output must be included Output s Spectrum Field Name must be unique and must follow the standard field naming conventions If any validation errors are detected then a validation error symbol will be displayed beside the table hovering the mouse cursor over a validation error symbol will show the error details Once all desired changes have been made click the OK button to apply the new changes or click Cancel to undo any changes made and close the Model Evaluator Options dialog Spectrum Technology Platform v9 0 SP3 Advanced Analytics Repository In this section Introduction to the Advanced Analytics Repository 12 e View Contents of Advanced Analytics Repository 12 e View Model DetailS ooooooommmmo 12 Add a Model to the Advanced Analytics Repository 13 e Delete a Model from the Advanced Analytics Repository 14 Introduction to the Advanced Analytics Repository Introduction to the Advanced Analytics Repository The Advanced Analytics Repository is where analytical models are saved for use with the Model Evaluator stage There is a web browser client for managing the models stored in the repository No
5. ch data by adding predicted or scored outputs to an existing data flow using models created by data insight teams using industry standard data modeling tools This may include models used to calculate the churn risk for existing customers or credit scoring models to determine a consumer s credit rating Advanced Analytics Module Components The Advanced Analytics Module consists of the following components Model Evaluator This stage can be used to evaluate any model stored in the Advanced Analytics Repository in the context of a data flow Advanced Analytics Repository This is the central repository for all models available to the Advanced Analytics Module Users can manage the repository via a web client 6 Spectrum Technology Platform v9 0 SP3 Model Evaluator In this section Introduction to the Model Evaluator 8 Deploying a 1066 eee 8 e Re configure Model Evaluator Settings 9 Introduction to the Model Evaluator Introduction to the Model Evaluator The Model Evaluator stage is capable of evaluating analytical models which have been published to the Advanced Analytics Repository in the context of a dataflow The evaluator operates on single data rows using the fields from each row as the inputs to the model User selected outputs from the model are written to the output channel Note For details of the supported model types and type mappings see Introducti
6. control will only allow page changes when the value chosen in step 2 for the number of models per page is less than the total number of models in the Advanced Analytics Repository Click on the Name column header to toggle sorting of listed models by the Name value in ascending descending order Alternatively click on the Type column header to toggle sorting of listed models by the Type value in ascending descending order Before any column headers are clicked the models by default are sorted by Name value in ascending order View Model Details This procedure describes how to view details about a model listed in the Advanced Analytics Repository to view it s description and details of the inputs and outputs for the model 12 1 Find the model you are looking for in the Repository page For details on how to browse the Advanced Analytics Repository see View Contents of Advanced Analytics Repository on page 12 Navigate to the model s detail page by either clicking on model s name in the Name column of the table or checking the model s checkbox and clicking the Model details button above the table The details page displays detailed information about the model Spectrum Technology Platform v9 0 SP3 Chapter 3 Advanced Analytics Repository Name the name of the model Type type of the model as described in Introduction to Supported Model Formats on page 16 Description a short text about the purpose of the model Inputs a table con
7. demarks of Pitney Bowes Software Inc All other marks and trademarks are property of their respective holders USPS Notices Pitney Bowes Inc holds a non exclusive license to publish and sell ZIP 4 databases on optical and magnetic media The following trademarks are owned by the United States Postal Service CASS CASS Certified DPV eLOT FASTforward First Class Mail Intelligent Mail LACS NCOAL PAVE PLANET Code Postal Service POSTNET Post Office RDI Suite United States Postal Service Standard Mail United States Post Office USPS ZIP Code and ZIP 4 This list is not exhaustive of the trademarks belonging to the Postal Service Pitney Bowes Inc is a non exclusive licensee of USPS for NCOAL 0 processing Prices for Pitney Bowes Software s products options and services are not established controlled or approved by USPS or United States Government When utilizing RDI data to determine parcel shipping costs the business decision on which parcel delivery company to use is not made by the USPS or United States Government Data Provider and Related Notices Data Products contained on this media and used within Pitney Bowes Software applications are protected by various trademarks and by one or more of the following copyrights O Copyright United States Postal Service All rights reserved O 2014 TomTom All rights reserved TomTom and the TomTom logo are registered trademarks of TomTom N V O Copyright NA
8. e name of the field as used in the Spectrum platform Model Field Type the type of the field as specified in the model definition Spectrum Field Type the Spectrum field type that is mapped to the model field type as described in QMML on page 16 and PMML on page 17 5 Select the Configuration tab Details about the model s Outputs are displayed in a table It contains information about all the output fields for the model Each row has information about an output field The model output field name will automatically be mapped to a valid Spectrum field name on publish see Add a Model to the Advanced Analytics Repository on page 13 Model Field Name the name of the field as specified in the model definition Spectrum Field Name the name of the field as used in the Spectrum platform Spectrum Technology Platform v9 0 SP3 Chapter 2 Model Evaluator Model Field Type the type of the field as specified in the model definition Spectrum Field Type the Spectrum field type that is mapped to the model field type as described in QMML on page 16 and PMML on page 17 Include a checkbox to specify whether to use this output 6 Optional Uncheck the Included column of any row in the Outputs table to exclude the output i e stop it from being written to the output channel At least one output must remain selected If all outputs are excluded then a validation error symbol will appear beside the Outputs table This means the current model config
9. e time not supported The Advanced Analytics Module currently supports PMML classification tree models only Classification Tree APMML classification tree model predicts membership of a categorical dependent variable from one or more independent variables Unsupported Features Classification trees with a missing value strategy of aggregateNodes or weightedConfidence are not supported Analytics Guide 17 PMML 18 Model Outputs Model Output Feature Target field predictedValue probability predictedDisplayValue transformedValue decision entityld Description If specified the target field will be available as an output feature this is a synonym for the predictedValue feature The categorical dependent variable that we are predicting membership of The statistical probability of the predicted value Multiple probability outputs can be specified in the model one for each predicted category or by rank The human readable value used to represent the predicted value from the model A value generated via a transformation expression applied to the predicted model output A value generated via an expression applied to the predicted model output resulting in a categorized value If present the ID of the tree node of the predicted result Spectrum Technology Platform v9 0 SP3 Notices 20 2014 Pitney Bowes Software Inc All rights reserved MapInfo and Group 1 Software are tra
10. el Miner Model 16 A Miner Model is any type of QMML model exported from Portrait Miner such as those generated from the decision tree scorecard cluster analysis or naive bayes modeling tools Unsupported Features All compiled QMML exported from Portrait Miner is supported Model Outputs Field Description lt dynamic fields gt A field is output for each QMML model output Spectrum Technology Platform v9 0 SP3 Chapter 4 Supported Model Types PMML Predictive Model Markup Language PMML is an XML based file format developed by the Data Mining Group to provide a way for applications to describe and exchange models produced by data mining and machine learning algorithms PMML files can be created and exported from many commercial and open source modeling tools Type Mapping PMML model inputs and outputs are automatically mapped to Spectrum Technology Platform field types All PMML models inputs and outputs must be of supported types PMML Field Type Spectrum Technology Platform Field Type string integer float double boolean date dateDaysSince 1960 dateDaysSince 1970 dateDaysSince 1980 dateDaysSince 0 time timeSeconds dateTime dateTimeSecondsSince 1960 dateTimeSecondsSince 1970 dateTimeSecondsSince 1980 dateTimeSecondsSince 0 Supported Models string integer float double boolean date date date date not supported time time date time date time date time dat
11. m Technology Platform v9 0 SP3 Supported Model Types In this section Introduction to Supported Model Formats Introduction to Supported Model Formats The Advanced Analytics Module supports the deployment of analytical models saved in both QMML and PMML file formats QMML models can be created and exported from Portrait Miner All types of analytical models and segmentations exported from Portrait Miner are supported by the Model Evaluator stage and can be deployed within a Spectrum Technology Platform dataflow PMML models can be created and exported from many commercial and open source modeling tools A list of the supported model types and their outputs is detailed in the PMML on page 17 section QMML QMML is a proprietary XML based file format used to represent model results generated from Portrait Miner All types of analytical models and segmentations exported from Portrait Miner are supported by the Model Evaluator stage and can be deployed within a Spectrum Technology Platform dataflow Type Mapping QMML model inputs and outputs are automatically mapped to Spectrum Technology Platform field types QMML Field Type Spectrum Technology Platform Field Type integer double string date time Supported Models All model types constructed within Portrait Miner including decision tree scorecard cluster analysis and naive bayes models are interpreted by the Advanced Analytics Module as a Miner Mod
12. on to Supported Model Formats on page 16 Deploying a Model This procedure describes how to configure the Model Evaluator stage to deploy an analytics model as part of a dataflow 1 Under Primary Stages Deployed Stages Advanced Analytics click the Model Evaluator stage and drag it onto the canvas placing it where you want on the dataflow and connecting it to input and output stages Double click the Model Evaluator stage to show the Model Evaluator Options dialog box By default the options dialog shows the details of the first model in the list of available models Click the Type Filter drop down and select the model type to filter by Only model types which are associated with at least one model in the Advanced Analytics Repository are listed Click the Model drop down and select the model to deploy The details of the chosen model are displayed in the dialog Model type the type of the selected model as described in Introduction to Supported Model Formats on page 16 Description a short text about the purpose of the model Inputs a table containing information about all the required input fields forthe model Each row contains information about an input field The model input field name will automatically be mapped to a valid Spectrum field name on publish see Add a Model to the Advanced Analytics Repository on page 13 Model Field Name the name of the field as specified in the model definition Spectrum Field Name th
13. rror Note White space at the start or end of a model s name is always removed This means if the name entered is already in use adding white space to the start or end of the name will not resolve the error Note Model names are case sensitive This means models can share the same name so long as they have different uppercase and lowercase characters For example models named my Model my model and mY ModEl can co exist simultaneously within the repository Note Name validation only detects conflicts against the contents of the Advanced Analytics Repository when the page was opened This means if the repository has changed since navigating to the Add new model page e g models have been added or deleted by other users then the page validation will not have these changes However on clicking the Publish button the new model will be re validated against the up to date version ofthe Advanced Analytics Repository 4 Write a description for the new model in to the text area labeled Description Models must have a description 5 Add a model definition by clicking the Choose file button and selecting a model stored in either QMML or PMML file formats Note The selected model must be less than 50MB in size Note When using an older browser that does not support file picking the XML model definition must be input into the text area labeled Upload model Note The model definition will not be validated until clicking the Publish but
14. s any new changes made to the model s configuration cannot be deployed step 5 Clicking Cancel will exit the Model Evaluator Options and allow the deleted model to be used within the dataflow as previously configured In order to apply any new changes to the stage configuration select a different non deleted model from the Model drop down list After applying a change of model further re configuration of the stage will no longer have the deleted model available in the Model drop down list 2 Optional From the Model tab select a different model from the Model drop down list to change which model is used within the dataflow Note On changing the selected model in the Model drop down list all configuration changes for the previously chosen model will be discarded Re selecting the model will result in the default output configuration for the model Clicking Cancel will undo any changes made since opening the Model Evaluator Options allowing any pending changes to be reverted Note On selecting a model if the model has been deleted since the Model Evaluator Options dialog was opened then a validation error symbol will be displayed beside the Model drop down list In this scenario the Inputs and Outputs will not be available and re configuration of the model settings will not be able to be applied until a non deleted model has been selected 3 Selectthe Configuration tab 4 Make any desired changes to the selected model s outputs configur
15. taining information about all the required fields for the model Each row contains a field s name and type as described in Introduction to Supported Model Formats on page 16 Outputs a table containing information about all the output fields produced by the model Each row contains a field s name and type as described in Introduction to Supported Model Formats on page 16 3 To navigate back to the Repository page click the Close button in the top right corner Add a Model to the Advanced Analytics Repository This procedure describes how to add a new model in to the Advanced Analytics Repository in order to make the model available for deployment in a Spectrum Technology Platform dataflow 1 Using the View Contents of Advanced Analytics Repository on page 12 procedure navigate to the Repository page 2 Click the Publish a model button located directly above the model table to navigate to the Add new model page The Publish a model button will be disabled if any models are selected in the model table s current page De select these models to re enable the Publish a model button 3 On the Add new model page there are three input fields Name Description and Upload model Write the name of the new model in to the text box labeled Name Model names must be unique If the model name inputted is already in use by another non deleted model stored in the Advanced Analytics Repository the text box will be highlighted with a validation e
16. te For details of the supported model types and type mappings see Introduction to Supported Model Formats on page 16 Accessing the Advanced Analytics Repository Web Browser Client In a web browser enter the URL http lt servername gt lt port gt analytics e g http localhost 8080 analytics to display the Login page Enter a valid Spectrum Technology Platform username and password in the browser dialog box displayed The browser will navigate to a page listing the existing Analytics models persisted in the repository View Contents of Advanced Analytics Repository This procedure describes how to view the list of models that are currently in the Advanced Analytics Repository 1 Log in to the Advanced Analytics Repository web browser client by entering a valid Spectrum Technology Platform username and password in the Login page On successful login the application will navigate to the Repository page showing a table listing the models in the Advanced Analytics Repository that are available for use by the Model Evaluator stage For details of how to log in via the Login page see Introduction to the Advanced Analytics Repository on page 12 Select the number of models to be shown per page from the Models per page drop down below the model table There is the option of displaying 10 20 or 50 models per table page The page control beside the Models per page drop down is used to navigate between the different pages The page
17. tion including without limitation the rights to use copy modify merge publish distribute and or sell copies of the Spectrum Technology Platform v9 0 SP3 Copyright Software and to permit persons to whom the Software is furnished to do so provided that the above copyright notice s and this permission notice appear in all copies of the Software and that both the above copyright notice s and this permission notice appear in supporting documentation THE SOFTWARE IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT OF THIRD PARTY RIGHTS IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS NOTICE BE LIABLE FOR ANY CLAIM OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE DATA OR PROFITS WHETHER IN AN ACTION OF CONTRACT NEGLIGENCE OR OTHER TORTIOUS ACTION ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE Except as contained in this notice the name of a copyright holder shall not be used in advertising or otherwise to promote the sale use or other dealings in this Software without prior written authorization of the copyright holder Analytics Guide 21
18. ton Analytics Guide 13 Delete a Model from the Advanced Analytics Repository 6 Click the Publish button to begin uploading the model to the Advanced Analytics Repository The Publish button will be disabled until the new model has a valid name and a description and model definition specified Wait as the Publish button s busy icon is shown No changes can be made to the new model s name description or model definition while the model is being published Once the model has been successfully published in to the Advanced Analytics Repository the page will navigate to the new model s Details page The Cancel button will be disabled while the model is being published Note Ifa problem is found while publishing the model e g the XML model definition is invalid or the model name is already in use then the model will not be published and an alert error will be displayed detailing the problem Click OK to close the error alert and correct the problem before attempting to publish the model again Delete a Model from the Advanced Analytics Repository This procedure describes how to delete models from the Advanced Analytics Repository Deleting a model from the repository will prevent the model from being deployed to any new Spectrum Technology Platform dataflows Deleting a model will not affect the runtime operation of any existing dataflows that are configured to use the model although no modifications can be made to the settings of a
19. uration is invalid and the model cannot be deployed The validation error symbol will remain visible until the error is corrected 7 Optional Click the Spectrum Field Name column of any row in the Outputs table to rename the fields as required No two outputs can share the same Spectrum Field Name and Spectrum Field Names must follow the standard Spectrum Technology Platform field naming conventions If any validation errors are detected then a validation error symbol is displayed beside the Outputs table hovering the mouse curser over a validation error symbol shows the error details Note Renaming an output s Spectrum Field Name only affects the specific instance of the stage it does not update the Advanced Analytics Repository 8 Finally click OK to save the chosen model and configuration Re configure Model Evaluator Settings This procedure describes how to re configure the Model Evaluator stage to re name or change the outputs that are generated from the Model Evaluator stage or to change the deployed model to be evaluated by the stage 1 Double click the Model Evaluator stage to show the Model Evaluator Options dialog box The options dialog shows the model which was previously selected configured and deployed Note Ifthe selected model has been deleted from the Advanced Analytics Repository prior to opening the Model Evaluator Options dialog a validation error symbol is displayed beside the Model drop down list This mean
Download Pdf Manuals
Related Search