JMP 13 Predictive and Specialized Modeling

JMP 13 Predictive and Specialized Modeling

by SAS Institute
Released September 2016
Publisher(s): SAS Institute
ISBN: 9781629605883

Read it now on the O’Reilly learning platform with a 10-day free trial.

O’Reilly members get unlimited access to books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.

Start your free trial

Book description

JMP 13 Predictive and Specialized Modeling provides details about modeling techniques such as partitioning, neural networks, nonlinear regression, and time series analysis. Topics include the Gaussian platform, which is useful in analyzing computer simulation experiments. The book also covers the Response Screening platform, which is useful in testing the effect of a predictor when you have many responses.

Table of contents

  1. Contents
  2. Learn about JMP
    1. Documentation and Additional Resources
    2. Formatting Conventions
    3. JMP Documentation
      1. JMP Documentation Library
      2. JMP Help
    4. Additional Resources for Learning JMP
      1. Tutorials
      2. Sample Data Tables
      3. Learn about Statistical and JSL Terms
      4. Learn JMP Tips and Tricks
      5. Tooltips
      6. JMP User Community
      7. JMPer Cable
      8. JMP Books by Users
      9. The JMP Starter Window
    5. Technical Support
  3. Introduction to Predictive and Specialized Modeling
    1. Overview of Modeling Techniques
  4. Modeling Utilities
    1. Exploring Data for Outliers, Missing Values, and Strong Predictors
    2. Explore Outliers Utility
      1. Example of the Explore Outliers Utility
      2. Launch the Explore Outliers Utility
        1. Quantile Range Outliers
        2. Robust Fit Outliers
        3. Multivariate Robust Outliers
        4. Multivariate k-Nearest Neighbor Outliers
      3. Explore Outliers Utility Options
      4. Additional Examples of the Explore Outliers Utility
        1. Multivariate k-Nearest Neighbor Outliers Example
    3. Explore Missing Values Utility
      1. Example of the Explore Missing Values Utility
      2. Examine Missing Values
      3. Impute Missing Values
      4. Launch the Explore Missing Values Utility
      5. The Missing Value Report
        1. Missing Value Report
        2. Missing Value Clustering
        3. Missing Value Snapshot
        4. Multivariate Normal Imputation
        5. Multivariate SVD Imputation
      6. Explore Missing Values Utility Options
    4. Make Validation Column Utility
      1. Example of the Make Validation Column Utility
      2. Launch the Make Validation Column Utility
        1. Make Validation Column Window
        2. Click Validation in a Platform Launch
  5. Neural Networks
    1. Fit Nonlinear Models Using Nodes and Layers
    2. Overview of Neural Networks
    3. Launch the Neural Platform
      1. The Neural Launch Window
      2. The Model Launch Control Panel
        1. Validation Method
        2. Hidden Layer Structure
        3. Boosting
        4. Fitting Options
    4. Model Reports
      1. Training and Validation Measures of Fit
      2. Confusion Statistics
    5. Model Options
    6. Example of a Neural Network
  6. Partition Models
    1. Use Decision Trees to Explore and Model Your Data
    2. Partition Platform Overview
    3. Example of the Partition Platform
    4. Launch the Partition Platform
    5. The Partition Report
      1. Control Buttons
      2. Report for Categorical Responses
        1. Partition Plot
        2. Summary Report
        3. Node Reports
      3. Report for Continuous Responses
        1. Partition Plot
        2. Summary Report
        3. Node Reports
    6. Partition Platform Options
      1. Show Fit Details
      2. Specify Profit Matrix
      3. Decision Matrix Report
      4. Informative Missing
      5. Actual by Predicted Plot
      6. ROC Curve
      7. Lift Curve
      8. Node Options
    7. Validation
      1. K-Fold Crossvalidation
        1. Crossvalidation Report
    8. Additional Examples of Partitioning
      1. Example of a Continuous Response
      2. Example of Informative Missing
      3. Example of Profit Matrix and Decision Matrix Report
    9. Statistical Details
      1. Responses and Factors
      2. Splitting Criterion
      3. Predicted Probabilities in Decision Tree and Bootstrap Forest
  7. Bootstrap Forest
    1. Fit a Model By Averaging Many Trees
    2. Bootstrap Forest Platform Overview
    3. Example of Bootstrap Forest with a Categorical Response
      1. Bootstrap Forest Model
      2. Missing Values
    4. Example of Bootstrap Forest with a Continuous Response
    5. Launch the Bootstrap Forest Platform
      1. Launch Window
      2. Specification Window
        1. Specification Panel
        2. Forest Panel
        3. Multiple Fits Panel
        4. Reproducibility Panel
    6. The Bootstrap Forest Report
      1. Model Validation-Set Summaries
      2. Specifications
      3. Overall Statistics
        1. Categorical Response
        2. Continuous Response
      4. Cumulative Validation
      5. Per-Tree Summaries
    7. Bootstrap Forest Platform Options
  8. Boosted Tree
    1. Fit Many Trees, Each Layered on the Previous
    2. Boosted Tree Platform Overview
    3. Example of Boosted Tree with a Categorical Response
    4. Example of Boosted Tree with a Continuous Response
    5. Launch the Boosted Tree Platform
      1. Launch Window
      2. Specification Window
        1. Boosting Panel
        2. Multiple Fits Panel
        3. Stochastic Boosting Panel
        4. Reproducibility Panel
        5. Early Stopping
    6. The Boosted Tree Report
      1. Model Validation - Set Summaries
      2. Specifications
      3. Overall Statistics
        1. Measures Report
        2. Confusion Matrix
        3. Decision Matrix
      4. Cumulative Validation
    7. Boosted Tree Platform Options
    8. Statistical Details for the Boosted Tree Platform
      1. Overfit Penalty
  9. K Nearest Neighbors
    1. Predict Response Values Using Nearby Observations
    2. K Nearest Neighbors Platform Overview
    3. Example of K Nearest Neighbors with Categorical Response
    4. Example of K Nearest Neighbors with Continuous Response
    5. Launch the K Nearest Neighbors Platform
    6. The K Nearest Neighbors Report
      1. Continuous Responses
      2. Categorical Responses
    7. K Nearest Neighbors Platform Options
  10. Naive Bayes
    1. Classify Observations Using Probabilistic Assumptions
    2. Naive Bayes Platform Overview
    3. Example of Naive Bayes
    4. Launch the Naive Bayes Platform
    5. The Naive Bayes Report
      1. Response Column Report
      2. Confusion Matrix Report
    6. Naive Bayes Platform Options
    7. Additional Example of Naive Bayes
    8. 'Statistical Details for the Naive Bayes Platform
      1. Algorithm
      2. Saved Probability Formulas
  11. Model Comparison
    1. Compare the Predictive Ability of Fitted Models
    2. Example of Model Comparison
    3. Launch the Model Comparison Platform
    4. The Model Comparison Report
    5. Model Comparison Platform Options
      1. Continuous and Categorical Responses
      2. Continuous Responses
      3. Categorical Responses
    6. Additional Example of Model Comparison
  12. Formula Depot
    1. Manage Models and Generate Scoring Code
    2. Formula Depot Platform Overview
    3. Example of Formula Depot
    4. Launch the Formula Depot Platform
      1. Platforms That Publish Prediction Formulas to the Formula Depot
    5. Formula Depot Platform Options
      1. Formula Depot Model Options
    6. Generating Scoring Code from the Formula Depot Platform
  13. Fit Curve
    1. Fit Built-In Nonlinear Models to Your Data
    2. Introduction to the Nonlinear Fit Curve Personality
    3. Example Using the Fit Curve Personality
    4. Launch the Nonlinear Platform
    5. The Fit Curve Report
      1. Initial Fit Curve Reports
        1. Model Comparison Report
        2. Model Reports
    6. Fit Curve Options
      1. Model Formulas
      2. Test Parallelism
      3. Compare Parameter Estimates
      4. Equivalence Test
  14. Nonlinear Regression
    1. Fit Custom Nonlinear Models to Your Data
    2. Example of Fitting a Custom Model
    3. Launch the Nonlinear Platform
    4. The Nonlinear Fit Report
    5. Nonlinear Platform Options
    6. Create a Formula Using the Model Library
      1. Customize the Nonlinear Model Library
    7. Additional Examples
      1. Example of Maximum Likelihood: Logistic Regression
      2. Example of a Probit Model with Binomial Errors: Numerical Derivatives
      3. Example of a Poisson Loss Function
      4. Example of Setting Parameter Limits
    8. Statistical Details
      1. Profile Likelihood Confidence Limits
      2. How Custom Loss Functions Work
      3. Notes Concerning Derivatives
      4. Notes on Effective Nonlinear Modeling
  15. Gaussian Process
    1. Fit Data Using Smoothing Models
    2. Example of Gaussian Process
    3. Launch the Gaussian Process Platform
    4. The Gaussian Process Report
      1. Actual by Predicted Plot
      2. Model Report
      3. Marginal Model Plots
    5. Gaussian Process Platform Options
    6. Additional Example of the Gaussian Process Platform
    7. Statistical Details for the Gaussian Process Platform
  16. Time Series Analysis
    1. Fit Time Series Models and Transfer Functions
    2. Launch the Platform
      1. Select Columns into Roles
      2. The Time Series Graph
    3. Time Series Commands
      1. Graph
      2. Autocorrelation and Partial Autocorrelation
        1. Statistical Details for Autocorrelation and Partial Autocorrelation
      3. Variogram
      4. AR Coefficients
      5. Spectral Density
      6. Save Spectral Density
      7. Difference
      8. Decomposition
        1. Remove Cycle
        2. X11
      9. Show Lag Plot
      10. Number of Forecast Periods
    4. Modeling Reports
      1. Model Comparison Table
        1. Model Comparison Options
      2. Model Summary Table
      3. Parameter Estimates Table
      4. Forecast Plot
      5. Residuals
      6. Iteration History
      7. Model Report Options
    5. ARIMA Model
    6. Seasonal ARIMA
    7. ARIMA Model Group
    8. Transfer Functions
      1. Report and Menu Structure
      2. Diagnostics
      3. Model Building
      4. Transfer Function Model
      5. Model Reports
      6. Model Comparison Table
      7. Fitting Notes
    9. Smoothing Models
      1. Simple Moving Average
      2. Smoothing Model Dialog
      3. Simple Exponential Smoothing
      4. Double (Brown) Exponential Smoothing
      5. Linear (Holt) Exponential Smoothing
      6. Damped-Trend Linear Exponential Smoothing
      7. Seasonal Exponential Smoothing
      8. Winters Method (Additive)
  17. Matched Pairs Analysis
    1. Compare Measurements on the Same Subject
    2. Overview of the Matched Pairs Platform
    3. Example of Comparing Matched Pairs
    4. Launch the Matched Pairs Platform
      1. Multiple Y Columns
    5. The Matched Pairs Report
      1. Difference Plot and Report
      2. Across Groups
    6. Matched Pairs Platform Options
    7. Example Comparing Matched Pairs across Groups
    8. Statistical Details for the Matched Pairs Platform
      1. Graphics for Matched Pairs
      2. Correlation of Responses
  18. Response Screening
    1. Test Many Responses in Large-Scale Data
    2. Response Screening Platform Overview
    3. Example of Response Screening
    4. Launch the Response Screening Platform
    5. The Response Screening Report
      1. FDR PValue Plot
      2. FDR LogWorth by Effect Size
      3. FDR LogWorth by RSquare
    6. The PValues Data Table
      1. PValues Data Table Columns
      2. Columns Added for Robust Option
      3. PValues Data Table Scripts
    7. Response Screening Platform Options
      1. Means Data Table
      2. Compare Means Data Table
    8. The Response Screening Personality in Fit Model
      1. Launch Response Screening in Fit Model
      2. The Fit Response Screening Report
      3. PValues Data Table
      4. Y Fits Data Table
    9. Additional Examples of Response Screening
      1. Example of Tests of Practical Significance and Equivalence
      2. Example of the MaxLogWorth Option
      3. Example of Robust Fit
      4. Response Screening Personality
    10. Statistical Details
      1. The False Discovery Rate
  19. Process Screening
    1. Screen Many Processes for Stability and Capability
    2. Process Screening Platform Overview
    3. Example of Process Screening
    4. Launch the Process Screening Platform
      1. Launch Window Roles
      2. Launch Window Options
      3. Limits Table
    5. The Process Screening Report
    6. Process Screening Platform Options
    7. Additional Example of Process Screening
    8. Statistical Details
      1. Scaling Factors for Using Medians to Estimate Sigma
  20. Predictor Screening
    1. Screen Many Predictors for Significant Effects
    2. Predictor Screening Platform Overview
    3. Example of Predictor Screening
    4. Launch the Predictor Screening Platform
    5. The Predictor Screening Report
    6. Predictor Screening Platform Options
  21. Association Analysis
    1. Perform Market Basket Analysis
    2. Association Analysis Platform Overview
    3. Example of the Association Analysis Platform
    4. Launch the Association Analysis Platform
    5. The Association Analysis Report
      1. Frequent Item Sets
      2. Rules
    6. Association Analysis Platform Options
      1. SVD
        1. Transaction Item Matrix
        2. Singular Value Decomposition
      2. SVD Report
        1. SVD Plots
        2. Singular Values
      3. Rotated SVD
        1. Topic Items
        2. Topic Scores
    7. Additional Example: SVD Analysis
    8. Statistical Details for the Association Analysis Platform
      1. Frequent Item Set Generation
      2. Association Analysis Performance Measures
        1. Support
        2. Confidence
        3. Lift
  22. References
  23. Index
    1. Specialized and Predictive Modeling

Product information

  • Title: JMP 13 Predictive and Specialized Modeling
  • Author(s): SAS Institute
  • Release date: September 2016
  • Publisher(s): SAS Institute
  • ISBN: 9781629605883