Foreword by Gareth James xix

Foreword by Ravi Bapna xxi

Preface to the Python Edition xxiii

Acknowledgments xxvii

Part I Preliminaries

Chapter 1 Introduction 3

1.1 What is Business Analytics? 3

1.2 What is Data Mining? 5

1.3 Data Mining and Related Terms 5

1.4 Big Data 6

1.5 Data Science 7

1.6 Why are There So Many Different Methods? 8

1.7 Terminology and Notation 9

1.8 Road Maps to This Book 11

Chapter 2 Overview of the Data Mining Process 15

2.1 Introduction 15

2.2 Core Ideas in Data Mining 16

2.3 The Steps in Data Mining 19

2.4 Preliminary Steps 21

2.5 Predictive Power and Overfitting 34

2.6 Building a Predictive Model 40

2.7 Using Python for Data Mining on a Local Machine 44

2.8 Automating Data Mining Solutions 45

2.9 Ethical Practice in Data Mining 47

Problems 56

Part II Data Exploration and Dimension Reduction

Chapter 3 Data Visualization 61

3.1 Introduction 61

3.2 Data Examples 64

3.3 Basic Charts: Bar Charts, Line Graphs, and Scatter Plots 65

3.4 Multidimensional Visualization 74

3.5 Specialized Visualizations 88

3.6 Summary: Major Visualizations and Operations, by Data Mining Goal 93

Problems 97

Chapter 4 Dimension Reduction 99

4.1 Introduction 100

4.2 Curse of Dimensionality 100

4.3 Practical Considerations 100

4.4 Data Summaries 102

4.5 Correlation Analysis 105

4.6 Reducing the Number of Categories in Categorical Variables 106

4.7 Converting a Categorical Variable to a Numerical Variable 108

4.8 Principal Components Analysis 108

4.9 Dimension Reduction Using Regression Models 119

4.10 Dimension Reduction Using Classification and Regression Trees 119

Problems 120

Part III Performance Evaluation

Chapter 5 Evaluating Predictive Performance 125

5.1 Introduction 126

5.2 Evaluating Predictive Performance 126

5.3 Judging Classifier Performance 131

5.4 Judging Ranking Performance 144

5.5 Oversampling 149

Problems 155

Part IV Prediction and Classification Methods

Chapter 6 Multiple Linear Regression 161

6.1 Introduction 162

6.2 Explanatory vs. Predictive Modeling 162

6.3 Estimating the Regression Equation and Prediction 164

6.4 Variable Selection in Linear Regression 169

Appendix: Using Statmodels 179

Problems 180

Chapter 7 k-Nearest Neighbors (kNN) 185

7.1 The k-NN Classifier (Categorical Outcome) 185

7.2 k-NN for a Numerical Outcome 193

7.3 Advantages and Shortcomings of k-NN Algorithms 195

Problems 197

Chapter 8 The Naive Bayes Classifier 199

8.1 Introduction 199

Example 1: Predicting Fraudulent Financial Reporting 201

8.2 Applying the Full (Exact) Bayesian Classifier 201

8.3 Advantages and Shortcomings of the Naive Bayes Classifier 210

Problems 214

Chapter 9 Classification and Regression Trees 217

9.1 Introduction 218

9.2 Classification Trees 220

9.3 Evaluating the Performance of a Classification Tree 228

9.4 Avoiding Overfitting 232

9.5 Classification Rules from Trees 238

9.6 Classification Trees for More Than Two Classes 239

9.7 Regression Trees 239

9.8 Improving Prediction: Random Forests and Boosted Trees 243

9.9 Advantages and Weaknesses of a Tree 246

Problems 248

Chapter 10 Logistic Regression 251

10.1 Introduction 252

10.2 The Logistic Regression Model 253

10.3 Example: Acceptance of Personal Loan 255

10.4 Evaluating Classification Performance 261

10.5 Logistic Regression for Multi-class Classification 264

10.6 Example of Complete Analysis: Predicting Delayed Flights 269

Appendix: Using Statmodels 278

Problems 280

Chapter 11 Neural Nets 283

11.1 Introduction 284

11.2 Concept and Structure of a Neural Network 284

11.3 Fitting a Network to Data 285

11.4 Required User Input 297

11.5 Exploring the Relationship Between Predictors and Outcome 299

11.6 Deep Learning 299

11.7 Advantages and Weaknesses of Neural Networks 305

Problems 306

Chapter 12 Discriminant Analysis 309

12.1 Introduction 310

12.2 Distance of a Record from a Class 311

12.3 Fisher’s Linear Classification Functions 314

12.4 Classification Performance of Discriminant Analysis 317

12.5 Prior Probabilities 318

12.6 Unequal Misclassification Costs 319

12.7 Classifying More Than Two Classes 319

12.8 Advantages and Weaknesses 322

Problems 324

Chapter 13 Combining Methods: Ensembles and Uplift Modeling 327

13.1 Ensembles 328

13.2 Uplift (Persuasion) Modeling 334

13.3 Summary 340

Problems 341

Part V Mining Relationships among Records

Chapter 14 Association Rules and Collaborative Filtering 345

14.1 Association Rules 346

14.2 Collaborative Filtering 357

14.3 Summary 368

Problems 370

Chapter 15 Cluster Analysis 375

15.1 Introduction 376

15.2 Measuring Distance Between Two Records 379

15.3 Measuring Distance Between Two Clusters 385

15.4 Hierarchical (Agglomerative) Clustering 387

15.5 Non-Hierarchical Clustering: The k-Means Algorithm 395

Problems 401

Part VI Forecasting Time Series

Chapter 16 Handling Time Series 407

16.1 Introduction 408

16.2 Descriptive vs. Predictive Modeling 409

16.3 Popular Forecasting Methods in Business 409

16.4 Time Series Components 410

16.5 Data-Partitioning and Performance Evaluation 415

Problems 419

Chapter 17 Regression-Based Forecasting 423

17.1 A Model with Trend 424

17.2 A Model with Seasonality 429

17.3 A Model with Trend and Seasonality 432

17.4 Autocorrelation and ARIMA Models 433

Problems 442

Chapter 18 Smoothing Methods 451

18.1 Introduction 452

18.2 Moving Average 452

18.3 Simple Exponential Smoothing 457

18.4 Advanced Exponential Smoothing 460

Problems 464

Part VII Data Analytics

Chapter 19 Social Network Analytics 473

19.1 Introduction 473

19.2 Directed vs. Undirected Networks 475

19.3 Visualizing and Analyzing Networks 476

19.4 Social Data Metrics and Taxonomy 480

19.5 Using Network Metrics in Prediction and Classification 485

19.6 Collecting Social Network Data with Python 491

19.7 Advantages and Disadvantages 491

Problems 494

Chapter 20 Text Mining 495

20.1 Introduction 496

20.2 The Tabular Representation of Text: Term-Document Matrix and “Bag-of-Words’’ 496

20.3 Bag-of-Words vs. Meaning Extraction at Document Level 497

20.4 Preprocessing the Text 498

20.5 Implementing Data Mining Methods 506

20.6 Example: Online Discussions on Autos and Electronics 506

20.7 Summary 510

Problems 511

Part VIII Cases

Chapter 21 Cases 515

21.1 Charles Book Club 515

21.2 German Credit 522

21.3 Tayko Software Cataloger 527

21.4 Political Persuasion 531

21.5 Taxi Cancellations 535

21.6 Segmenting Consumers of Bath Soap 537

21.7 Direct-Mail Fundraising 541

21.8 Catalog Cross-Selling 544

21.9 Time Series Case: Forecasting Public Transportation Demand 546

References 549

Data Files Used in the Book 551

Python Utilities Functions 555

Index 565