| |
| |
| Preface | |
| |
| |
| |
| What Is Data Mining? | |
| |
| |
| |
| Big Data | |
| |
| |
| |
| The Data Warehouse | |
| |
| |
| |
| Timelines | |
| |
| |
| |
| Types of Data-Mining Problems | |
| |
| |
| |
| The Pedigree of Data Mining | |
| |
| |
| |
| Databases | |
| |
| |
| |
| Statistics | |
| |
| |
| |
| Machine Learning | |
| |
| |
| |
| Is Big Better? | |
| |
| |
| |
| Strong Statistical Evaluation | |
| |
| |
| |
| More Intensive Search | |
| |
| |
| |
| More Controlled Experiments | |
| |
| |
| |
| Is Big Necessary? | |
| |
| |
| |
| The Tasks of Predictive Data Mining | |
| |
| |
| |
| Data Preparation | |
| |
| |
| |
| Data Reduction | |
| |
| |
| |
| Data Modeling and Prediction | |
| |
| |
| |
| Case and Solution Analyses | |
| |
| |
| |
| Data Mining: Art or Science? | |
| |
| |
| |
| An Overview of the Book | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Statistical Evaluation for Big Data | |
| |
| |
| |
| The Idealized Model | |
| |
| |
| |
| Classical Statistical Comparison and Evaluation | |
| |
| |
| |
| It's Big but Is It Biased? | |
| |
| |
| |
| Objective Versus Survey Data | |
| |
| |
| |
| Significance and Predictive Value | |
| |
| |
| |
| Too Many Comparisons? | |
| |
| |
| |
| Classical Types of Statistical Prediction | |
| |
| |
| |
| Predicting True-or-False: Classification | |
| |
| |
| |
| Error Rates | |
| |
| |
| |
| Forecasting Numbers: Regression | |
| |
| |
| |
| Distance Measures | |
| |
| |
| |
| Measuring Predictive Performance | |
| |
| |
| |
| Independent Testing | |
| |
| |
| |
| Random Training and Testing | |
| |
| |
| |
| How Accurate Is the Error Estimate? | |
| |
| |
| |
| Comparing Results for Error Measures | |
| |
| |
| |
| Ideal or Real-World Sampling? | |
| |
| |
| |
| Training and Testing from Different Time Periods | |
| |
| |
| |
| Too Much Searching and Testing? | |
| |
| |
| |
| Why Are Errors Made? | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Preparing the Data | |
| |
| |
| |
| A Standard Form | |
| |
| |
| |
| Standard Measurements | |
| |
| |
| |
| Goals | |
| |
| |
| |
| Data Transformations | |
| |
| |
| |
| Normalizations | |
| |
| |
| |
| Data Smoothing | |
| |
| |
| |
| Differences and Ratios | |
| |
| |
| |
| Missing Data | |
| |
| |
| |
| Time-Dependent Data | |
| |
| |
| |
| Time Series | |
| |
| |
| |
| Composing Features from Time Series | |
| |
| |
| |
| Current Values | |
| |
| |
| |
| Moving Averages | |
| |
| |
| |
| Trends | |
| |
| |
| |
| Seasonal Adjustments | |
| |
| |
| |
| Hybrid Time-Dependent Applications | |
| |
| |
| |
| Multivariate Time Series | |
| |
| |
| |
| Classification and Time Series | |
| |
| |
| |
| Standard Cases with Time-Series Attributes | |
| |
| |
| |
| Text Mining | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Data Reduction | |
| |
| |
| |
| Selecting the Best Features | |
| |
| |
| |
| Feature Selection from Means and Variances | |
| |
| |
| |
| Independent Features | |
| |
| |
| |
| Distance-Based Optimal Feature Selection | |
| |
| |
| |
| Heuristic Feature Selection | |
| |
| |
| |
| Principal Components | |
| |
| |
| |
| Feature Selection by Decision Trees | |
| |
| |
| |
| How Many Measured Values? | |
| |
| |
| |
| Reducing and Smoothing Values | |
| |
| |
| |
| Rounding | |
| |
| |
| |
| K-Means Clustering | |
| |
| |
| |
| Class Entropy | |
| |
| |
| |
| How Many Cases? | |
| |
| |
| |
| A Single Sample | |
| |
| |
| |
| Incremental Samples | |
| |
| |
| |
| Average Samples | |
| |
| |
| |
| Specialized Case-Reduction Techniques | |
| |
| |
| |
| Sequential Sampling over Time | |
| |
| |
| |
| Strategic Sampling of Key Events | |
| |
| |
| |
| Adjusting Prevalence | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Looking for Solutions | |
| |
| |
| |
| Overview | |
| |
| |
| |
| Math Solutions | |
| |
| |
| |
| Linear Scoring | |
| |
| |
| |
| Nonlinear Scoring: Neural Nets | |
| |
| |
| |
| Advanced Statistical Methods | |
| |
| |
| |
| Distance Solutions | |
| |
| |
| |
| Logic Solutions | |
| |
| |
| |
| Decision Trees | |
| |
| |
| |
| Decision Rules | |
| |
| |
| |
| What Do the Answers Mean? | |
| |
| |
| |
| Is It Safe to Edit Solutions? | |
| |
| |
| |
| Which Solution Is Preferable? | |
| |
| |
| |
| Combining Different Answers | |
| |
| |
| |
| Multiple Prediction Methods | |
| |
| |
| |
| Multiple Samples | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| What's Best for Data Reduction and Mining? | |
| |
| |
| |
| Let's Analyze Some Real Data | |
| |
| |
| |
| The Experimental Methods | |
| |
| |
| |
| The Empirical Results | |
| |
| |
| |
| Significance Testing | |
| |
| |
| |
| So What Did We Learn? | |
| |
| |
| |
| Feature Selection | |
| |
| |
| |
| Value Reduction | |
| |
| |
| |
| Subsampling or All Cases? | |
| |
| |
| |
| Graphical Trend Analysis | |
| |
| |
| |
| Incremental Case Analysis | |
| |
| |
| |
| Incremental Complexity Analysis | |
| |
| |
| |
| Maximum Data Reduction | |
| |
| |
| |
| Are There Winners and Losers in Performance? | |
| |
| |
| |
| Getting the Best Results | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Art or Science? Case Studies in Data Mining | |
| |
| |
| |
| Why These Case Studies? | |
| |
| |
| |
| A Summary of Tasks for Predictive Data Mining | |
| |
| |
| |
| A Checklist for Data Preparation | |
| |
| |
| |
| A Checklist for Data Reduction | |
| |
| |
| |
| A Checklist for Data Modeling and Prediction | |
| |
| |
| |
| A Checklist for Case and Solution Analyses | |
| |
| |
| |
| The Case Studies | |
| |
| |
| |
| Transaction Processing | |
| |
| |
| |
| Text Mining | |
| |
| |
| |
| Outcomes Analysis | |
| |
| |
| |
| Process Control | |
| |
| |
| |
| Marketing and User Profiling | |
| |
| |
| |
| Exploratory Analysis | |
| |
| |
| |
| Looking Ahead | |
| |
| |
| |
| Bibliographic and Historical Remarks | |
| |
| |
| |
| Data-Miner Software Kit | |
| |
| |
| References | |
| |
| |
| Author Index | |
| |
| |
| Subject Index | |