book

JMP 13 Predictive and Specialized Modeling

Name: JMP 13 Predictive and Specialized Modeling
Author: SAS Institute
ISBN: 9781629605883

by SAS Institute

September 2016

Intermediate to advanced

366 pages

9h 24m

English

SAS Institute

Read now

Unlock full access

Contents
Learn about JMP
Documentation and Additional Resources
Formatting Conventions
JMP DocumentationJMP Documentation LibraryJMP HelpAdditional Resources for Learning JMPTutorialsSample Data TablesLearn about Statistical and JSL TermsLearn JMP Tips and TricksTooltipsJMP User CommunityJMPer CableJMP Books by UsersThe JMP Starter WindowTechnical Support
Introduction to Predictive and Specialized Modeling
Overview of Modeling Techniques
Modeling Utilities
Exploring Data for Outliers, Missing Values, and Strong Predictors
Explore Outliers Utility
Example of the Explore Outliers UtilityLaunch the Explore Outliers UtilityQuantile Range OutliersRobust Fit OutliersMultivariate Robust OutliersMultivariate k-Nearest Neighbor OutliersExplore Outliers Utility OptionsAdditional Examples of the Explore Outliers UtilityMultivariate k-Nearest Neighbor Outliers ExampleExplore Missing Values UtilityExample of the Explore Missing Values UtilityExamine Missing ValuesImpute Missing ValuesLaunch the Explore Missing Values UtilityThe Missing Value ReportMissing Value ReportMissing Value ClusteringMissing Value SnapshotMultivariate Normal ImputationMultivariate SVD ImputationExplore Missing Values Utility OptionsMake Validation Column UtilityExample of the Make Validation Column UtilityLaunch the Make Validation Column UtilityMake Validation Column WindowClick Validation in a Platform Launch
Neural Networks
Fit Nonlinear Models Using Nodes and Layers
Overview of Neural Networks
Launch the Neural PlatformThe Neural Launch WindowThe Model Launch Control PanelValidation MethodHidden Layer StructureBoostingFitting OptionsModel ReportsTraining and Validation Measures of FitConfusion StatisticsModel OptionsExample of a Neural Network
Partition Models
Use Decision Trees to Explore and Model Your Data
Partition Platform Overview
Example of the Partition PlatformLaunch the Partition PlatformThe Partition ReportControl ButtonsReport for Categorical ResponsesPartition PlotSummary ReportNode ReportsReport for Continuous ResponsesPartition PlotSummary ReportNode ReportsPartition Platform OptionsShow Fit DetailsSpecify Profit MatrixDecision Matrix ReportInformative MissingActual by Predicted PlotROC CurveLift CurveNode OptionsValidationK-Fold CrossvalidationCrossvalidation ReportAdditional Examples of PartitioningExample of a Continuous ResponseExample of Informative MissingExample of Profit Matrix and Decision Matrix ReportStatistical DetailsResponses and FactorsSplitting CriterionPredicted Probabilities in Decision Tree and Bootstrap Forest

Bootstrap Forest
Fit a Model By Averaging Many Trees
Bootstrap Forest Platform Overview
Example of Bootstrap Forest with a Categorical ResponseBootstrap Forest ModelMissing ValuesExample of Bootstrap Forest with a Continuous ResponseLaunch the Bootstrap Forest PlatformLaunch WindowSpecification WindowSpecification PanelForest PanelMultiple Fits PanelReproducibility PanelThe Bootstrap Forest ReportModel Validation-Set SummariesSpecificationsOverall StatisticsCategorical ResponseContinuous ResponseCumulative ValidationPer-Tree SummariesBootstrap Forest Platform Options
Boosted Tree
Fit Many Trees, Each Layered on the Previous
Boosted Tree Platform Overview
Example of Boosted Tree with a Categorical ResponseExample of Boosted Tree with a Continuous ResponseLaunch the Boosted Tree PlatformLaunch WindowSpecification WindowBoosting PanelMultiple Fits PanelStochastic Boosting PanelReproducibility PanelEarly StoppingThe Boosted Tree ReportModel Validation - Set SummariesSpecificationsOverall StatisticsMeasures ReportConfusion MatrixDecision MatrixCumulative ValidationBoosted Tree Platform OptionsStatistical Details for the Boosted Tree PlatformOverfit Penalty
K Nearest Neighbors
Predict Response Values Using Nearby Observations
K Nearest Neighbors Platform Overview
Example of K Nearest Neighbors with Categorical ResponseExample of K Nearest Neighbors with Continuous ResponseLaunch the K Nearest Neighbors PlatformThe K Nearest Neighbors ReportContinuous ResponsesCategorical ResponsesK Nearest Neighbors Platform Options
Naive Bayes
Classify Observations Using Probabilistic Assumptions
Naive Bayes Platform Overview
Example of Naive BayesLaunch the Naive Bayes PlatformThe Naive Bayes ReportResponse Column ReportConfusion Matrix ReportNaive Bayes Platform OptionsAdditional Example of Naive Bayes'Statistical Details for the Naive Bayes PlatformAlgorithmSaved Probability Formulas
Model Comparison
Compare the Predictive Ability of Fitted Models
Example of Model Comparison
Launch the Model Comparison PlatformThe Model Comparison ReportModel Comparison Platform OptionsContinuous and Categorical ResponsesContinuous ResponsesCategorical ResponsesAdditional Example of Model Comparison
Formula Depot
Manage Models and Generate Scoring Code
Formula Depot Platform Overview
Example of Formula DepotLaunch the Formula Depot PlatformPlatforms That Publish Prediction Formulas to the Formula DepotFormula Depot Platform OptionsFormula Depot Model OptionsGenerating Scoring Code from the Formula Depot Platform
Fit Curve
Fit Built-In Nonlinear Models to Your Data
Introduction to the Nonlinear Fit Curve Personality
Example Using the Fit Curve PersonalityLaunch the Nonlinear PlatformThe Fit Curve ReportInitial Fit Curve ReportsModel Comparison ReportModel ReportsFit Curve OptionsModel FormulasTest ParallelismCompare Parameter EstimatesEquivalence Test
Nonlinear Regression
Fit Custom Nonlinear Models to Your Data
Example of Fitting a Custom Model
Launch the Nonlinear PlatformThe Nonlinear Fit ReportNonlinear Platform OptionsCreate a Formula Using the Model LibraryCustomize the Nonlinear Model LibraryAdditional ExamplesExample of Maximum Likelihood: Logistic RegressionExample of a Probit Model with Binomial Errors: Numerical DerivativesExample of a Poisson Loss FunctionExample of Setting Parameter LimitsStatistical DetailsProfile Likelihood Confidence LimitsHow Custom Loss Functions WorkNotes Concerning DerivativesNotes on Effective Nonlinear Modeling
Gaussian Process
Fit Data Using Smoothing Models
Example of Gaussian Process
Launch the Gaussian Process PlatformThe Gaussian Process ReportActual by Predicted PlotModel ReportMarginal Model PlotsGaussian Process Platform OptionsAdditional Example of the Gaussian Process PlatformStatistical Details for the Gaussian Process Platform
Time Series Analysis
Fit Time Series Models and Transfer Functions
Launch the Platform
Select Columns into RolesThe Time Series GraphTime Series CommandsGraphAutocorrelation and Partial AutocorrelationStatistical Details for Autocorrelation and Partial AutocorrelationVariogramAR CoefficientsSpectral DensitySave Spectral DensityDifferenceDecompositionRemove CycleX11Show Lag PlotNumber of Forecast PeriodsModeling ReportsModel Comparison TableModel Comparison OptionsModel Summary TableParameter Estimates TableForecast PlotResidualsIteration HistoryModel Report OptionsARIMA ModelSeasonal ARIMAARIMA Model GroupTransfer FunctionsReport and Menu StructureDiagnosticsModel BuildingTransfer Function ModelModel ReportsModel Comparison TableFitting NotesSmoothing ModelsSimple Moving AverageSmoothing Model DialogSimple Exponential SmoothingDouble (Brown) Exponential SmoothingLinear (Holt) Exponential SmoothingDamped-Trend Linear Exponential SmoothingSeasonal Exponential SmoothingWinters Method (Additive)
Matched Pairs Analysis
Compare Measurements on the Same Subject
Overview of the Matched Pairs Platform
Example of Comparing Matched PairsLaunch the Matched Pairs PlatformMultiple Y ColumnsThe Matched Pairs ReportDifference Plot and ReportAcross GroupsMatched Pairs Platform OptionsExample Comparing Matched Pairs across GroupsStatistical Details for the Matched Pairs PlatformGraphics for Matched PairsCorrelation of Responses
Response Screening
Test Many Responses in Large-Scale Data
Response Screening Platform Overview
Example of Response ScreeningLaunch the Response Screening PlatformThe Response Screening ReportFDR PValue PlotFDR LogWorth by Effect SizeFDR LogWorth by RSquareThe PValues Data TablePValues Data Table ColumnsColumns Added for Robust OptionPValues Data Table ScriptsResponse Screening Platform OptionsMeans Data TableCompare Means Data TableThe Response Screening Personality in Fit ModelLaunch Response Screening in Fit ModelThe Fit Response Screening ReportPValues Data TableY Fits Data TableAdditional Examples of Response ScreeningExample of Tests of Practical Significance and EquivalenceExample of the MaxLogWorth OptionExample of Robust FitResponse Screening PersonalityStatistical DetailsThe False Discovery Rate
Process Screening
Screen Many Processes for Stability and Capability
Process Screening Platform Overview
Example of Process ScreeningLaunch the Process Screening PlatformLaunch Window RolesLaunch Window OptionsLimits TableThe Process Screening ReportProcess Screening Platform OptionsAdditional Example of Process ScreeningStatistical DetailsScaling Factors for Using Medians to Estimate Sigma
Predictor Screening
Screen Many Predictors for Significant Effects
Predictor Screening Platform Overview
Example of Predictor ScreeningLaunch the Predictor Screening PlatformThe Predictor Screening ReportPredictor Screening Platform Options
Association Analysis
Perform Market Basket Analysis
Association Analysis Platform Overview
Example of the Association Analysis PlatformLaunch the Association Analysis PlatformThe Association Analysis ReportFrequent Item SetsRulesAssociation Analysis Platform OptionsSVDTransaction Item MatrixSingular Value DecompositionSVD ReportSVD PlotsSingular ValuesRotated SVDTopic ItemsTopic ScoresAdditional Example: SVD AnalysisStatistical Details for the Association Analysis PlatformFrequent Item Set GenerationAssociation Analysis Performance MeasuresSupportConfidenceLift
References
Index
Specialized and Predictive Modeling

Content preview from JMP 13 Predictive and Specialized Modeling

Chapter 7

Boosted Tree

Fit Many Trees, Each Layered on the Previous

The Boosted Tree platform is available only in JMP Pro.

Boosting is the process of building a large, additive decision tree by fitting a sequence of smaller decision trees, called layers. The tree at each layer consists of a small number of splits. The tree is fit based on the residuals of the previous layers, which allows each layer to correct the fit for bad fitting data from the previous layers. The final prediction for an observation is the sum of the predicted residuals for that observation over all the layers.

Figure 7.1 Example of Boosted Tree Layers

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.

Read now

Unlock full access

More than 5,000 organizations count on O’Reilly

O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.

Julian F.

Head of Cybersecurity

I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.

Addison B.

Field Engineer

I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.

Amir M.

Data Platform Tech Lead

I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.

Mark W.

Embedded Software Engineer

JMP 13 Predictive and Specialized Modeling, Second Edition, 2nd Edition

Publisher Resources

ISBN: 9781629605883

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills

JMP 13 Predictive and Specialized Modeling

by SAS Institute

Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.