Foreword xvii

Preface to the second edition xix

Preface to the first edition xxi

Acknowledgments xxiii

Part I PRELIMINARIES

Chapter 1 Introduction 3

1.1 What Is Data Mining? 3

1.2 Where Is Data Mining Used? 4

1.3 Origins of Data Mining 4

1.4 Rapid Growth of Data Mining 5

1.5 Why Are There So Many Different Methods? 6

1.6 Terminology and Notation 7

1.7 Road Maps to This Book 9

Chapter 2 Overview of the Data Mining Process 12

2.1 Introduction 12

2.2 Core Ideas in Data Mining 13

2.3 Supervised and Unsupervised Learning 15

2.4 Steps in Data Mining 15

2.5 Preliminary Steps 17

2.6 Building a Model: Example with Linear Regression 27

2.7 Using Excel for Data Mining 34

Part II DATA EXPLORATION AND DIMENSION REDUCTION

Chapter 3 Data Visualization 43

3.1 Uses of Data Visualization 43

3.2 Data Examples 45

3.3 Basic Charts: Bar Charts, Line Graphs, and Scatterplots 45

3.4 Multidimensional Visualization 52

3.5 Specialized Visualizations 63

3.6 Summary ofMajor Visualizations and Operations, According to Data Mining Goal 67

Chapter 4 Dimension Reduction 71

4.1 Introduction 71

4.2 Practical Considerations 72

4.3 Data Summaries 73

4.4 Correlation Analysis . 76

4.5 Reducing the Number of Categories in Categorical Variables 76

4.6 Converting a Categorical Variable to a Numerical Variable 78

4.7 Principal Components Analysis 78

4.8 Dimension Reduction Using Regression Models 87

4.9 Dimension Reduction Using Classification and Regression Trees 88

Part III PERFORMANCE EVALUATION

Chapter 5 Evaluating Classification and Predictive Performance 93

5.1 Introduction 93

5.2 Judging Classification Performance 94

5.3 Evaluating Predictive Performance 115

Part IV PREDICTION AND CLASSIFICATION METHODS

Chapter 6 Multiple Linear Regression 121

6.1 Introduction 121

6.2 Explanatory versus Predictive Modeling 122

6.3 Estimating the Regression Equation and Prediction 123

6.4 Variable Selection in Linear Regression 127

Chapter 7 k-Nearest Neighbors (k-NN) 137

7.1 k-NN Classifier (Categorical Outcome) 137

7.2 k-NN for a Numerical Response 142

7.3 Advantages and Shortcomings of k-NN Algorithms 144

Chapter 8 Naive Bayes 148

8.1 Introduction 148

8.2 Applying the Full (Exact) Bayesian Classifier 150

8.3 Advantages and Shortcomings of the Naive Bayes Classifier 159

Chapter 9 Classification and Regression Trees 164

9.1 Introduction 164

9.2 Classification Trees 166

9.3 Measures of Impurity 169

9.4 Evaluating the Performance of a Classification Tree 173

9.5 Avoiding Overfitting 179

9.6 Classification Rules from Trees 183

9.7 Classification Trees for More Than Two Classes 185

9.8 RegressionTrees 185

9.9 Advantages, Weaknesses, and Extensions 187

Chapter 10 Logistic Regression 192

10.1 Introduction 192

10.2 Logistic Regression Model 194

10.3 Evaluating Classification Performance 202

10.4 Example of Complete Analysis: Predicting Delayed Flights 206

10.5 Appendix: Logistic Regression for Profiling 211

Chapter 11 Neural Nets 222

11.1 Introduction 222

11.2 Concept and Structure of a Neural Network 223

11.3 Fitting a Network to Data 223

11.4 Required User Input 237

11.5 Exploring the Relationship Between Predictors andResponse 239

11.6 Advantages and Weaknesses of Neural Networks 239

Chapter 12 Discriminant Analysis 243

12.1 Introduction 243

12.2 Distance of an Observation from a Class 246

12.3 Fisher’s Linear Classification Functions 247

12.4 Classification Performance of Discriminant Analysis 251

12.5 Prior Probabilities 252

12.6 Unequal Misclassification Costs 252

12.7 Classifying More Than Two Classes 253

Part V MINING RELATIONSHIPS AMONG RECORDS

Chapter 13 Association Rules 263

13.1 Introduction 263

13.2 Discovering Association Rules in Transaction Databases 263

13.3 Generating Candidate Rules 265

13.4 Selecting Strong Rules 267

13.5 Summary 275

Chapter 14 Cluster Analysis 279

14.1 Introduction 279

14.2 Measuring Distance Between Two Records 283

14.3 Measuring Distance Between Two Clusters 287

14.4 Hierarchical (Agglomerative) Clustering 290

14.5 Nonhierarchical Clustering: The k-Means Algorithm 295

Part VI FORECASTING TIME SERIES

Chapter 15 Handling Time Series 305

15.1 Introduction 305

15.2 Explanatory versus Predictive Modeling 306

15.3 Popular Forecasting Methods in Business 307

15.4 Time Series Components 308

15.5 Data Partitioning 312

Chapter 16 Regression-Based Forecasting 317

16.1 Model with Trend 317

16.2 Model with Seasonality 322

16.3 Model with Trend and Seasonality 324

16.4 Autocorrelation and ARIMA Models 324

Chapter 17 Smoothing Methods 344

17.1 Introduction 344

17.2 MovingAverage 345

17.3 Simple Exponential Smoothing 350

Part VII CASES

Chapter 18 Cases 367

18.1 Charles Book Club 367

18.2 German Credit 375

18.3 Tayko Software Cataloger 379

18.4 Segmenting Consumers of Bath Soap 383

18.5 Direct-MailFundraising 387

18.6 Catalog Cross Selling 389

18.7 Predicting Bankruptcy 390

18.8 Time Series Case: Forecasting Public Transportation Demand 393

References 397

Index 399