Cover
Wiley & SAS Business Series
Title Page
Foreword
Acknowledgments
CHAPTER 1: A Value-Centric Perspective Towards Analytics
1. INTRODUCTION
2. PROFIT-DRIVEN BUSINESS ANALYTICS
3. ANALYTICS PROCESS MODEL
4. ANALYTICAL MODEL EVALUATION
5. ANALYTICS TEAM
6. CONCLUSION
7. REVIEW QUESTIONS
8. REFERENCES
CHAPTER 2: Analytical Techniques
1. INTRODUCTION
2. DATA PREPROCESSING
3. TYPES OF ANALYTICS
4. PREDICTIVE ANALYTICS
5. ENSEMBLE METHODS
6. EVALUATING PREDICTIVE MODELS
7. DESCRIPTIVE ANALYTICS
8. SURVIVAL ANALYSIS
9. SOCIAL NETWORK ANALYTICS
10. CONCLUSION
11. REVIEW QUESTIONS
12. NOTES
13. REFERENCES
CHAPTER 3: Business Applications
1. INTRODUCTION
2. MARKETING ANALYTICS
3. FRAUD ANALYTICS
4. CREDIT RISK ANALYTICS
5. HR ANALYTICS
6. CONCLUSION
7. REVIEW QUESTIONS
8. NOTE
9. REFERENCES
CHAPTER 4: Uplift Modeling
1. INTRODUCTION
2. EXPERIMENTAL DESIGN, DATA COLLECTION, AND DATA PREPROCESSING
3. UPLIFT MODELING METHODS
4. EVALUATION OF UPLIFT MODELS
5. PRACTICAL GUIDELINES
6. CONCLUSION
7. REVIEW QUESTIONS
8. NOTE
9. REFERENCES
CHAPTER 5: Profit-Driven Analytical Techniques
1. INTRODUCTION
2. PROFIT-DRIVEN PREDICTIVE ANALYTICS
3. COST-SENSITIVE CLASSIFICATION
4. COST-SENSITIVE REGRESSION
5. COST-SENSITIVE LEARNING FOR REGRESSION
6. PROFIT-DRIVEN DESCRIPTIVE ANALYTICS
7. CONCLUSION
8. REVIEW QUESTIONS
9. NOTES
10. REFERENCES
CHAPTER 6: Profit-Driven Model Evaluation and Implementation
1. INTRODUCTION
2. PROFIT-DRIVEN EVALUATION OF CLASSIFICATION MODELS
3. PROFIT-DRIVEN EVALUATION OF REGRESSION MODELS
4. CONCLUSION
5. REVIEW QUESTIONS
6. NOTES
7. REFERENCES
CHAPTER 7: Economic Impact
1. INTRODUCTION
2. ECONOMIC VALUE OF BIG DATA AND ANALYTICS
3. KEY ECONOMIC CONSIDERATIONS
4. IMPROVING THE ROI OF BIG DATA AND ANALYTICS
5. CONCLUSION
6. REVIEW QUESTIONS
7. NOTES
8. REFERENCES
About the Authors
Index
End User License Agreement

List of Tables

Chapter 1
1. Table 1.1 Categories of Analytics from a Task-Oriented Perspective
2. Table 1.2 Example Datasets and Predictive Analytical Models
3. Table 1.3 Example Datasets and Descriptive Analytical Models
4. Table 1.4 Structured Dataset
5. Table 1.5 Examples of Business Decisions Matching Analytics
6. Table 1.6 Outline of the Book
7. Table 1.7 Key Characteristics of Successful Business Analytics Models
Chapter 2
1. Table 2.1 Missing Values in a Dataset
2. Table 2.2 Dataset for Linear Regression
3. Table 2.3 Example Classification Dataset
4. Table 2.4 Reference Values for Variable Significance
5. Table 2.5 Example Dataset for Performance Calculation
6. Table 2.6 The Confusion Matrix
7. Table 2.7 Receiver Operating Characteristic (ROC) Analysis
8. Table 2.8 Example Transaction Dataset
9. Table 2.9 The Lift Measure
10. Table 2.10 Example Transaction Dataset (left) and Sequential Dataset (right) for Sequence Rule Mining
11. Table 2.11 Matrix Representation of a Social Network
12. Table 2.12 Network Centrality Measures
13. Table 2.13 Centrality measures for the Kite network
Chapter 3
1. Table 3.1 K-Means Clustering Sample Output
2. Table 3.2 Event Log of Customer Activities
3. Table 3.3 Example User-Item Matrix
4. Table 3.4 Example Call Detail Record Dataset for Fraud Detection
Chapter 4
1. Table 4.1 Overview of Model and Campaign Effect Measurement
2. Table 4.2 Example Model and Campaign Effect Measurement
3. Table 4.3 Dataset Including Treatment Dummy Variable t, Predictor Variables x_i and Target Variable y
4. Table 4.4 Relabeled Dataset of Table 4.3 Following Lai's Method
5. Table 4.5 Relabeled Dataset of Table 4.3 Following the Generalized Lai Method
Chapter 5
1. Table 5.1 Confusion Matrix
2. Table 5.2 Cost Matrix for a Binary Classification Problem
3. Table 5.3 Example Cost Matrix for the German Credit Data
4. Table 5.4 Cost Matrix for an Ordinal Classification Problem
5. Table 5.5 Simplified Cost Matrix
6. Table 5.6 Structured Overview of the Cost-Sensitive Classification Approaches Discussed in This Chapter
7. Table 5.7 Overview of Sampling Based Ensemble Approaches for Cost-Sensitive Learning
8. Table 5.8 Overview of Weighting Approaches for Cost-Sensitive Learning
9. Table 5.9 Overview of Cost-Sensitive Decision-Tree Approaches
10. Table 5.10 Overview of Cost-Sensitive Boosting Based Approaches
11. Table 5.11 Example of the MetaCost Approach
12. Table 5.12 Elaborated Example of Algorithm 5.1
13. Table 5.13 Average CLV and Fraction of Observations per Segment
14. Table 5.14 Average CLV, RFM Variables and Cost of Service for the Observations in the Five K-means Clusters
Chapter 6
1. Table 6.1 Cost Matrix for a Binary Classification Problem
2. Table 6.2 Cost Matrix for the German Credit Data Example Dataset
3. Table 6.3 Cutoff Points, Accuracies, and Costs. The cutoff with the best accuracy is 0.90, and the one with the lowest average misclassification cost is 0.55
4. Table 6.4 Accuracy and AUC for the Five Candidate Models
5. Table 6.5 AUC and AUC of the Convex Hull for the Five Candidate Models
6. Table 6.6 H-Measure for the Five Candidate Models. Model 3 is the selected one, not Model 5
7. Table 6.7 Confusion Matrix for a Binary Classification Problem
8. Table 6.8 EMP and Selected Fraction per Model
9. Table 6.9 Benefits, Costs, and Profits for the Test Set. Model 3 gives the better performance, as expected
10. Table 6.10 Observation-Dependent Cost Matrix for Credit Scoring
11. Table 6.11 Overview of Standard Regression Evaluation Measures
12. Table 6.12 Data for REC Curve
13. Table 6.13 Data for REC Surface
Chapter 7
1. Table 7.1 Example Costs for Calculating Total Cost of Ownership (TCO)
2. Table 7.2 Data Quality Dimensions (Wang et al. 1996)

List of Illustrations

Chapter 1
1. Figure 1.1 The analytics process model.
2. Figure 1.2 Profile of a data scientist.
Chapter 2
1. Figure 2.1 Aggregating normalized data tables into a non-normalized data table.
2. Figure 2.2 Example dataset showing an ellipse rotated in 45 degrees.
3. Figure 2.3 PCA of the simulated data.
4. Figure 2.4 OLS regression.
5. Figure 2.5 Bounding function for logistic regression.
6. Figure 2.6 Linear decision boundary of logistic regression.
7. Figure 2.7 Calculating the p-value with a Student's t-distribution.
8. Figure 2.8 Variable subsets for four variables x₁, x₂, x₃, and x₄.
9. Figure 2.9 Example decision tree.
10. Figure 2.10 Example datasets for calculating impurity.
11. Figure 2.11 Entropy versus Gini.
12. Figure 2.12 Calculating the entropy for age split.
13. Figure 2.13 Using a validation set to stop growing a decision tree.
14. Figure 2.14 Decision boundary of a decision tree.
15. Figure 2.15 Example regression tree for predicting the fraud percentage.
16. Figure 2.16 Neural network representation of logistic regression.
17. Figure 2.17 A Multilayer Perceptron (MLP) neural network.
18. Figure 2.18 Local versus global minima.
19. Figure 2.19 Using a validation set for stopping neural network training.
20. Figure 2.20 Training and test set split-up for performance estimation.
21. Figure 2.21 Cross-validation for performance measurement.
22. Figure 2.22 Bootstrapping.
23. Figure 2.23 Receiver operating characteristic curve.
24. Figure 2.24 The lift curve.
25. Figure 2.25 The cumulative accuracy profile (CAP).
26. Figure 2.26 Calculating the accuracy ratio.
27. Figure 2.27 Scatter plot.
28. Figure 2.28 Hierarchical versus nonhierarchical clustering techniques.
29. Figure 2.29 Divisive versus agglomerative hierarchical clustering.
30. Figure 2.30 Euclidean versus Manhattan distance.
31. Figure 2.31 Calculating distances between clusters.
32. Figure 2.32 Example for clustering birds. The numbers indicate the clustering steps.
33. Figure 2.33 Dendrogram for birds example. The red line indicates the optimal clustering.
34. Figure 2.34 Scree plot for clustering.
35. Figure 2.35 Rectangular versus hexagonal SOM grid.
36. Figure 2.36 Clustering countries using SOMs.
37. Figure 2.37 Component plane for literacy.
38. Figure 2.38 Component plane for political rights.
39. Figure 2.39 Example of right censoring for churn prediction.
40. Figure 2.40 Example of a discrete event time distribution.
41. Figure 2.41 Cumulative distribution and survival function for the event time distribution in Figure 2.40.
42. Figure 2.42 Sample hazard shapes.
43. Figure 2.43 Kaplan Meier example.
44. Figure 2.44 Exponential event time distribution, with cumulative distribution and hazard function.
45. Figure 2.45 Weibull distributions.
46. Figure 2.46 The proportional hazards model.
47. Figure 2.47 Sociogram representation of a social network.
48. Figure 2.48 The Kite network.
49. Figure 2.49 Example social network for relational neighbor classifier.
50. Figure 2.50 Example social network for probabilistic relational neighbor classifier.
51. Figure 2.51 Relational logistic regression.
52. Figure 2.52 Example of featurization with features describing target behavior of neighbors.
53. Figure 2.53 Example of featurization with features describing local node behavior of neighbors.
Chapter 3
1. Figure 3.1 Constructing an RFM score (independent sorting).
2. Figure 3.2 Constructing an RFM score (dependent sorting).
3. Figure 3.3 Cluster profiling using histograms.
4. Figure 3.4 Using decision trees for clustering interpretation.
5. Figure 3.5 Example Markov chain (Pfeifer and Carraway 2000).
6. Figure 3.6 Customer journey in a mortgage sales process.
7. Figure 3.7 Example social network for fraud detection.
8. Figure 3.8 Multilevel credit risk model architecture.
9. Figure 3.9 Example employee network.
Chapter 4
1. Figure 4.1 Four types of customers identified as a function of purchasing behavior when treated or not treated.
2. Figure 4.2 Experimental design to collect the required data for uplift modeling, allowing the selection of a model base for the campaign.
3. Figure 4.3 Categorization of customers based on whether a customer was treated and whether the customer responded.
4. Figure 4.4 (a) High uplift but low number of observations in the left child node versus (b) lower uplift but applicable to a higher number of observations.
5. Figure 4.5 Illustration of Gain_U calculation.
6. Figure 4.6 Response rate by decile graph for both treatment and control groups (upper panel) and uplift by deciles graph (lower panel).
7. Figure 4.7 Response rate curve for the perfect uplift model, plotting the response rates for the treatment and control groups ranked according to estimated uplift.
8. Figure 4.8 Uplift by decile curve of an accurate uplift model.
9. Figure 4.9 Cumulative incremental gains charts or Qini curves for two uplift models and the baseline model.
Chapter 5
1. Figure 5.1 Oversampling the fraudsters.
2. Figure 5.2 Undersampling the nonfraudsters.
3. Figure 5.3 Synthetic minority oversampling technique (SMOTE).
4. Figure 5.4 Quadratic cost versus true cost as a function of the prediction error.
5. Figure 5.5 Linlin cost function C_linlin in function of the prediction error e.
6. Figure 5.6 Average misprediction cost in function of the adjustment δ.
7. Figure 5.7 Customer lifetime value distribution for a dataset containing 1,000 customers.
8. Figure 5.8 Three-cut strategy for CLV segmentation.
9. Figure 5.9 Three-group customer segmentation.
10. Figure 5.10 Plot of the first two principal components following a K-means clustering of the CLV example dataset.
11. Figure 5.11 Density maps of CLV example dataset SOMs of different sizes.
12. Figure 5.12 Distance map of the CLV example dataset SOM.
13. Figure 5.13 Codebook vector graph of the CLV example dataset SOM.
14. Figure 5.14 Heatmaps for the variables in the CLV example dataset.
15. Figure 5.15 Dendrogram plot of the hierarchical clustering procedure using single linkage.
16. Figure 5.16 Codebook vector graph with clustering limits superimposed.
Chapter 6
1. Figure 6.1 Illustration of the two-cutoff point strategy.
2. Figure 6.2 Receiver operating characteristic curve.
3. Figure 6.3 Convex hull for a nonconcave ROC curve.
4. Figure 6.4 Beta distribution for different values of the parameters α and β.
5. Figure 6.5 ROC curves for five credit risk models.
6. Figure 6.6 Customer churn management process. Adapted from Verbraken et al. (2013).
7. Figure 6.7 LGD histogram and percentage of observations per score group.
8. Figure 6.8 Regression error characteristic (REC) curve.
9. Figure 6.9 Regression error characteristic surface following the data in Table 6.13.
Chapter 7
1. Figure 7.1 ROI of big data and analytics.

Wiley & SAS Business Series

The Wiley & SAS Business Series presents books that help senior-level managers with their critical management decisions.

Titles in the Wiley & SAS Business Series include:

Analytics: The Agile Way by Phil Simon
Analytics in a Big Data World: The Essential Guide to Data Science and its Applications by Bart Baesens
A Practical Guide to Analytics for Governments: Using Big Data for Good by Marie Lowman
Bank Fraud: Using Technology to Combat Losses by Revathi Subramanian
Big Data Analytics: Turning Big Data into Big Money by Frank Ohlhorst
Big Data, Big Innovation: Enabling Competitive Differentiation through Business Analytics by Evan Stubbs
Business Analytics for Customer Intelligence by Gert Laursen
Business Intelligence Applied: Implementing an Effective Information and Communications Technology Infrastructure by Michael Gendron
Business Intelligence and the Cloud: Strategic Implementation Guide by Michael S. Gendron
Business Transformation: A Roadmap for Maximizing Organizational Insights by Aiman Zeid
Connecting Organizational Silos: Taking Knowledge Flow Management to the Next Level with Social Media by Frank Leistner
Data-Driven Healthcare: How Analytics and BI are Transforming the Industry by Laura Madsen
Delivering Business Analytics: Practical Guidelines for Best Practice by Evan Stubbs
Demand-Driven Forecasting: A Structured Approach to Forecasting, Second Edition by Charles Chase
Demand-Driven Inventory Optimization and Replenishment: Creating a More Efficient Supply Chain by Robert A. Davis
Developing Human Capital: Using Analytics to Plan and Optimize Your Learning and Development Investments by Gene Pease, Barbara Beresford, and Lew Walker
The Executive's Guide to Enterprise Social Media Strategy: How Social Networks Are Radically Transforming Your Business by David Thomas and Mike Barlow
Economic and Business Forecasting: Analyzing and Interpreting Econometric Results by John Silvia, Azhar Iqbal, Kaylyn Swankoski, Sarah Watt, and Sam Bullard
Economic Modeling in the Post Great Recession Era: Incomplete Data, Imperfect Markets by John Silvia, Azhar Iqbal, and Sarah Watt House
Enhance Oil & Gas Exploration with Data-Driven Geophysical and Petrophysical Models by Keith Holdaway and Duncan Irving
Foreign Currency Financial Reporting from Euros to Yen to Yuan: A Guide to Fundamental Concepts and Practical Applications by Robert Rowan
Fraud Analytics Using Descriptive, Predictive, and Social Network Techniques: A Guide to Data Science for Fraud Detection by Bart Baesens, Veronique Van Vlasselaer, and Wouter Verbeke
Harness Oil and Gas Big Data with Analytics: Optimize Exploration and Production with Data-Driven Models by Keith Holdaway
Health Analytics: Gaining the Insights to Transform Health Care by Jason Burke
Heuristics in Analytics: A Practical Perspective of What Influences Our Analytical World by Carlos Andre Reis Pinheiro and Fiona McNeill
Human Capital Analytics: How to Harness the Potential of Your Organization's Greatest Asset by Gene Pease, Boyce Byerly, and Jac Fitz-enz
Implement, Improve and Expand Your Statewide Longitudinal Data System: Creating a Culture of Data in Education by Jamie McQuiggan and Armistead Sapp
Intelligent Credit Scoring: Building and Implementing Better Credit Risk Scorecards, Second Edition, by Naeem Siddiqi
Killer Analytics: Top 20 Metrics Missing from your Balance Sheet by Mark Brown
Machine Learning for Marketers: Hold the Math by Jim Sterne
On-Camera Coach: Tools and Techniques for Business Professionals in a Video-Driven World by Karin Reed
Predictive Analytics for Human Resources by Jac Fitz-enz and John Mattox II
Predictive Business Analytics: Forward-Looking Capabilities to Improve Business Performance by Lawrence Maisel and Gary Cokins
Profit Driven Business Analytics: A Practitioner's Guide to Transforming Big Data into Added Value by Wouter Verbeke, Cristian Bravo, and Bart Baesens
Retail Analytics: The Secret Weapon by Emmett Cox
Social Network Analysis in Telecommunications by Carlos Andre Reis Pinheiro
Statistical Thinking: Improving Business Performance, Second Edition by Roger W. Hoerl and Ronald D. Snee
Strategies in Biomedical Data Science: Driving Force for Innovation by Jay Etchings
Style & Statistic: The Art of Retail Analytics by Brittany Bullard
Taming the Big Data Tidal Wave: Finding Opportunities in Huge Data Streams with Advanced Analytics by Bill Franks
Too Big to Ignore: The Business Case for Big Data by Phil Simon
The Analytic Hospitality Executive by Kelly A. McGuire
The Value of Business Analytics: Identifying the Path to Profitability by Evan Stubbs
The Visual Organization: Data Visualization, Big Data, and the Quest for Better Decisions by Phil Simon
Using Big Data Analytics: Turning Big Data into Big Money by Jared Dean
Win with Advanced Business Analytics: Creating Business Value from Your Data by Jean Paul Isson and Jesse Harriott

For more information on any of the above titles, please visit www.wiley.com.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 646-8600, or on the Web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at www.wiley.com/go/permissions.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993, or fax (317) 572-4002.

Wiley publishes in a variety of print and electronic formats and by print-on-demand. Some material included with standard print versions of this book may not be included in e-books or in print-on-demand. If this book refers to media such as a CD or DVD that is not included in the version you purchased, you may download this material at http://booksupport.wiley.com. For more information about Wiley products, visit www.wiley.com.

Library of Congress Cataloging-in-Publication Data is Available:

ISBN 9781119286554 (Hardcover)

ISBN 9781119286998 (ePDF)

ISBN 9781119286981 (ePub)

Cover Design: Wiley

Profit-Driven Business Analytics

A Practitioner’s Guide to Transforming Big Data into Added Value

Wiley & SAS Business Series

Foreword

Acknowledgments