book

Predictive Modeling with SAS Enterprise Miner, 2nd Edition

Name: Predictive Modeling with SAS Enterprise Miner, 2nd Edition
Author: Kattamuri S. Sarma
ISBN: 9781607648185

by Kattamuri S. Sarma

December 2013

Intermediate to advanced

500 pages

13h 56m

English

SAS Institute

Read now

Unlock full access

Preface
About This Book
About The Author
Acknowledgments
Chapter 1: Research Strategy
1.1 Introduction1.2 Measurement Scales for Variables1.3 Defining the Target1.3.1 Predicting Response to Direct Mail1.3.2 Predicting Risk in the Auto Insurance Industry1.3.3 Predicting Rate Sensitivity of Bank Deposit Products1.3.4 Predicting Customer Attrition1.3.5 Predicting a Nominal Categorical (Unordered Polychotomous)Target1.4 Sources of Modeling Data1.4.1 Comparability between the Sample and Target Universe1.4.2 Observation Weights1.5 Pre-Processing the Data1.5.1 Data Cleaning Before Launching SAS Enterprise Miner1.5.2 Data Cleaning After Launching SAS Enterprise Miner1.6 Alternative Modeling Strategies1.6.1 Regression with a Moderate Number of Input Variables1.6.2 Regression with a Large Number of Input Variables1.7 Notes
Chapter 2: Getting Started with Predictive Modeling
2.1 Introduction2.2 Opening SAS Enterprise Miner 12.12.3 Creating a New Project in SAS Enterprise Miner 12.12.4 The SAS Enterprise Miner Window2.5 Creating a SAS Data Source2.6 Creating a Process Flow Diagram2.7 Sample Nodes2.7.1 Input Data Node2.7.2 Data Partition Node2.7.3 Filter Node2.7.4 File Import Node2.7.5 Time Series Node2.7.6 Merge Node2.7.7 Append Node2.8 Tools for Initial Data Exploration2.8.1 Stat Explore Node2.8.2 MultiPlot Node2.8.3 Graph Explore Node2.8.4 Variable Clustering Node2.8.5 Cluster Node2.8.6 Variable Selection Node2.9 Tools for Data Modification2.9.1 Drop Node2.9.2 Replacement Node2.9.3 Impute Node2.9.4 Interactive Binning Node2.9.5 Principal Components Node2.9.6 Transform Variables Node2.10 Utility Nodes2.10.1 SAS Code Node2.11 Appendix to Chapter 22.11.1 The Type, the Measurement Scale, and the Number of Levels of a Variable2.11.2 Eigenvalues, Eigenvectors, and Principal Components2.11.3 Cramer’s V2.11.4 Calculation of Chi-Square Statistic and Cramer’s V for a Continuous Input2.12 Exercises
Chapter 3: Variable Selection and Transformation of Variables
3.1 Introduction3.2 Variable Selection3.2.1 Continuous Target with Numeric Interval-scaled Inputs (Case 1)3.2.2 Continuous Target with Nominal-Categorical Inputs (Case 2)3.2.3 Binary Target with Numeric Interval-scaled Inputs (Case 3)3.2.4 Binary Target with Nominal-scaled Categorical Inputs (Case 4)3.3 Variable Selection Using the Variable Clustering Node3.3.1 Selection of the Best Variable from Each Cluster3.3.2 Selecting the Cluster Components3.4 Variable Selection Using the Decision Tree Node3.5 Transformation of Variables3.5.1 Transform Variables Node3.5.2 Transformation before Variable Selection3.5.3 Transformation after Variable Selection3.5.4 Passing More Than One Type of Transformation for Each Interval Input to the Next Node3.5.5 Saving and Exporting the Code Generated by the Transform Variables Node3.6 Summary3.7 Appendix to Chapter 33.7.1 Changing the Measurement Scale of a Variable in a Data Source3.7.2 SAS Code for Comparing Grouped Categorical Variables with the Ungrouped VariablesExercisesNote
Chapter 4: Building Decision Tree Models to Predict Response and Risk
4.1 Introduction4.2 An Overview of the Tree Methodology in SAS Enterprise Miner4.2.1 Decision Trees4.2.2 Decision Tree Models4.2.3 Decision Tree Models vs. Logistic Regression Models4.2.4 Applying the Decision Tree Model to Prospect Data4.2.5 Calculation of the Worth of a Tree4.2.6 Roles of the Training and Validation Data in the Development of a Decision Tree4.2.7 Regression Tree4.3 Development of the Tree in SAS Enterprise Miner4.3.1 Growing an Initial Tree4.3.2 P-value Adjustment Options4.3.3 Controlling Tree Growth: Stopping Rules4.3.4 Pruning: Selecting the Right-Sized Tree Using Validation Data4.3.5 Step-by-Step Illustration of Growing and Pruning a Tree4.3.6 Average Profit vs. Total Profit for Comparing Trees of Different Sizes4.3.7 Accuracy /Misclassification Criterion in Selecting the Right-sized Tree (Classification of Records and Nodes by Maximizing Accuracy)4.3.8 Assessment of a Tree or Sub-tree Using Average Square Error4.3.9 Selection of the Right-sized Tree4.4 A Decision Tree Model to Predict Response to Direct Marketing4.4.1 Testing Model Performance with a Test Data Set4.4.2 Applying the Decision Tree Model to Score a Data Set4.5 Developing a Regression Tree Model to Predict Risk4.5.1 Summary of the Regression Tree Model to Predict Risk4.6 Developing Decision Trees Interactively4.6.1 Interactively Modifying an Existing Decision Tree4.6.2 Growing a Tree Interactively Starting from the Root Node4.6.3 Developing the Maximal Tree in Interactive Mode4.7 Summary4.8 Appendix to Chapter 44.8.1 Pearson’s Chi-Square Test4.8.2 Adjusting the Predicted Probabilities for Over-sampling4.8.3 Expected Profits Using Unadjusted Probabilities4.8.4 Expected Profits Using Adjusted Probabilities4.9 Exercises
Chapter 5: Neural Network Models to Predict Response and Risk
5.1 Introduction5.1.1 Target Variables for the Models5.1.2 Neural Network Node Details5.2 A General Example of a Neural Network Model5.2.1 Input Layer5.2.2 Hidden Layers5.2.3 Output Layer or Target Layer5.2.4 Activation Function of the Output Layer5.3 Estimation of Weights in a Neural Network Model5.4 A Neural Network Model to Predict Response5.4.1 Setting the Neural Network Node Properties5.4.2 Assessing the Predictive Performance of the Estimated Model5.4.3 Receiver Operating Characteristic (ROC) Charts5.4.4 How Did the Neural Network Node Pick the Optimum Weights for This Model?5.4.5 Scoring a Data Set Using the Neural Network Model5.4.6 Score Code5.5 A Neural Network Model to Predict Loss Frequency in Auto Insurance5.5.1 Loss Frequency as an Ordinal Target5.5.3 Classification of Risks for Rate Setting in Auto Insurance with Predicted Probabilities5.6 Alternative Specifications of the Neural Networks5.6.1 A Multilayer Perceptron (MLP) Neural Network5.6.2 A Radial Basis Function (RBF) Neural Network5.7 Comparison of Alternative Built-in Architectures of the Neural Network Node5.7.1 Multilayer Perceptron (MLP) Network5.7.2 Ordinary Radial Basis Function with Equal Heights and Widths (ORBFEQ)5.7.3 Ordinary Radial Basis Function with Equal Heights and Unequal Widths (ORBFUN)5.7.4 Normalized Radial Basis Function with Equal Widths and Heights (NRBFEQ)5.7.5 Normalized Radial Basis Function with Equal Heights and Unequal Widths (NRBFEH)5.7.6 Normalized Radial Basis Function with Equal Widths and Unequal Heights (NRBFEW)5.7.7 Normalized Radial Basis Function with Equal Volumes (NRBFEV)5.7.8 Normalized Radial Basis Function with Unequal Widths and Heights (NRBFUN)5.7.9 User-Specified Architectures5.8 AutoNeural Node5.9 DMNeural Node5.10 Dmine Regression Node5.11 Comparing the Models Generated by DMNeural, AutoNeural, and Dmine Regression Nodes5.12 Summary5.13 Appendix to Chapter 55.14 Exercises
Chapter 6: Regression Models
6.1 Introduction6.2 What Types of Models Can Be Developed Using the Regression Node?6.2.1 Models with a Binary Target6.2.2 Models with an Ordinal Target6.2.3 Models with a Nominal (Unordered) Target6.2.4 Models with Continuous Targets6.3 An Overview of Some Properties of the Regression Node6.3.1 Regression Type Property6.3.2 Link Function Property6.3.3 Selection Model Property6.3.4 Selection Criterion Property6.4 Business Applications6.4.1 Logistic Regression for Predicting Response to a Mail Campaign6.4.2 Regression for a Continuous Target6.5 Summary6.6 Appendix to Chapter 66.6 Exercises

Chapter 7: Comparison and Combination of Different Models
7.1 Introduction7.2 Models for Binary Targets: An Example of Predicting Attrition7.2.1 Logistic Regression for Predicting Attrition7.2.2 Decision Tree Model for Predicting Attrition7.2.3 A Neural Network Model for Predicting Attrition7.3 Models for Ordinal Targets: An Example of Predicting the Risk of Accident Risk7.3.1 Lift Charts and Capture Rates for Models with Ordinal Targets7.3.2 Logistic Regression with Proportional Odds for Predicting Risk in Auto Insurance7.3.3 Decision Tree Model for Predicting Risk in Auto Insurance7.3.4 Neural Network Model for Predicting Risk in Auto Insurance7.4 Comparison of All Three Accident Risk Models7.5 Boosting and Combining Predictive Models7.5.1 Gradient Boosting7.5.2 Stochastic Gradient Boosting7.5.3 An Illustration of Boosting Using the Gradient Boosting Node7.5.4 The Ensemble Node7.5.5 Comparing the Gradient Boosting and Ensemble Methods of Combining Models7.6 Appendix to Chapter 77.6.1 Least Squares Loss7.6.2 Least Absolute Deviation Loss7.6.3 Huber-M Loss7.6.4 Logit Loss7.7 Exercises
Chapter 8: Customer Profitability
8.1 Introduction8.2 Acquisition Cost8.3 Cost of Default8.4 Revenue8.5 Profit8.6 The Optimum Cut-off Point8.7 Alternative Scenarios of Response and Risk8.8 Customer Lifetime Value8.9 Suggestions for Extending Results
Chapter 9: Introduction to Predictive Modeling with Textual Data
9.1 Introduction9.1.1 Quantifying Textual Data: A Simplified Example9.1.2 Dimension Reduction and Latent Semantic Indexing9.1.3 Summary of the Steps in Quantifying Textual Information9.2 Retrieving Documents from the World Wide Web9.2.1 The %TMFILTER Macro9.3 Creating a SAS Data Set from Text Files9.4 The Text Import Node9.5 Creating a Data Source for Text Mining9.6 Text Parsing Node9.7 Text Filter Node9.7.1 Frequency Weighting9.7.2 Term Weighting9.7.3 Adjusted Frequencies9.7.4 Frequency Weighting Methods9.7.5 Term Weighting Methods9.8 Text Topic Node9.8.1 Developing a Predictive Equation Using the Output Data Set Created by the Text Topic Node9.9 Text Cluster Node9.9.1 Hierarchical Clustering9.9.2 Expectation-Maximization (EM) Clustering9.9.3 Using the Text Cluster Node9.10 Exercises
Index

Overview

Learn the theory behind and methods for predictive modeling using SAS Enterprise Miner.

Learn how to produce predictive models and prepare presentation-quality graphics in record time with Predictive Modeling with SAS Enterprise Miner: Practical Solutions for Business Applications, Second Edition.

If you are a graduate student, researcher, or statistician interested in predictive modeling; a data mining expert who wants to learn SAS Enterprise Miner; or a business analyst looking for an introduction to predictive modeling using SAS Enterprise Miner, you'll be able to develop predictive models quickly and effectively using the theory and examples presented in this book.

Author Kattamuri Sarma offers the theory behind, programming steps for, and examples of predictive modeling with SAS Enterprise Miner, along with exercises at the end of each chapter. You'll gain a comprehensive awareness of how to find solutions for your business needs. This second edition features expanded coverage of the SAS Enterprise Miner nodes, now including File Import, Time Series, Variable Clustering, Cluster, Interactive Binning, Principal Components, AutoNeural, DMNeural, Dmine Regression, Gradient Boosting, Ensemble, and Text Mining.

Develop predictive models quickly, learn how to test numerous models and compare the results, gain an in-depth understanding of predictive models and multivariate methods, and discover how to do in-depth analysis. Do it all with Predictive Modeling with SAS Enterprise Miner.

This book is part of the SAS Press program.

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

Predictive Modeling with SAS Enterprise Miner, 3rd Edition

Publisher Resources

ISBN: 9781607648185

Cloud Computing

Data Engineering

Data Science

AI & ML

Programming Languages

Software Architecture

IT/Ops

Security

Design

Business

Soft Skills