Intro; Preface; Contents; Acronyms; 1 Introduction to KDD and Data Science; 1.1 Introduction; 1.2 A Definition of Data Science; 1.3 The Data Science Process; 1.3.1 Selection of the Data; 1.3.2 Data Preprocessing; 1.3.2.1 Why Is Preprocessing Required?; 1.3.3 Stages of the Data Preprocessing Phase; 1.3.3.1 Selection of Data; 1.3.3.2 Exploration of Data; 1.3.3.3 Transformation of Data; 1.4 Standard Data Science Problems; 1.4.1 Descriptive Problems; 1.4.2 Predictive Problems; 1.5 Classical Data Mining Techniques; 1.6 Non-standard Data Science Problems; 1.6.1 Derivative Problems

1.6.1.1 Imbalanced Learning1.6.1.2 Multi-instance Learning; 1.6.1.3 Multi-label Classification; 1.6.1.4 Data Stream Learning; 1.6.2 Hybrid Problems; 1.6.2.1 Semi-supervised Learning; 1.6.2.2 Subgroup Discovery; 1.6.2.3 Ordinal Classification/Regression; 1.6.2.4 Transfer Learning; References; 2 Foundations on Imbalanced Classification; 2.1 Formal Description; 2.2 Applications; 2.2.1 Engineering; 2.2.2 Information Technology; 2.2.3 Bioinformatics; 2.2.4 Medicine; 2.2.4.1 Quality Control; 2.2.4.2 Medical Diagnosis; 2.2.4.3 Medical Prognosis; 2.2.5 Business Management; 2.2.6 Security

2.2.7 Education2.3 Case Studies on Imbalanced Classification; References; 3 Performance Measures; 3.1 Introduction; 3.2 Nominal Class Predictions; 3.3 Scoring Predictions; 3.4 Probabilistic Predictions; 3.5 Summarizing Comments; References; 4 Cost-Sensitive Learning; 4.1 Introduction; 4.2 Obtaining the Cost Matrix; 4.3 MetaCost; 4.4 Cost-Sensitive Decision Trees; 4.4.1 Direct Approach with Cost-Sensitive Splitting; 4.4.2 Meta-learning Approach with Instance Weighting; 4.5 Other Cost-Sensitive Classifiers; 4.5.1 Support Vector Machines; 4.5.2 Artificial Neural Networks; 4.5.3 Nearest Neighbors

4.6 Hybrid Cost-Sensitive Approaches4.7 Summarizing Comments; References; 5 Data Level Preprocessing Methods; 5.1 Introduction; 5.2 Undersampling and Oversampling Basics; 5.3 Advanced Undersampling Techniques; 5.3.1 Evolutionary Undersampling; 5.3.1.1 ACOSampling; 5.3.1.2 IPADE-ID; 5.3.1.3 CBEUS: Cluster-Based Evolutionary Undersampling; 5.3.2 Undersampling by Cleaning Data; 5.3.2.1 Weighted Sampling; 5.3.2.2 IHT: Instance Hardness Threshold; 5.3.2.3 Hybrid Undersampling; 5.3.3 Ensemble Based Undersampling; 5.3.3.1 IRUS: Inverse Random Undersampling

5.3.3.2 OligoIS: Oligarchic Instance Selection5.3.4 Clustering Based Undersampling; 5.3.4.1 ClusterOSS; 5.3.4.2 DSUS: Diversified Sensitivity Undersampling; 5.4 Synthetic Minority Oversampling TEchnique (SMOTE); 5.5 Extensions of SMOTE; 5.5.1 Borderline-SMOTE; 5.5.2 Adjusting the Direction of the Synthetic Minority ClasS Examples: ADOMS; 5.5.3 ADASYN: Adaptive Synthetic Sampling Approach; Input; Procedure; 5.5.4 ROSE: Random Oversampling Examples; 5.5.5 Safe-Level-SMOTE; 5.5.6 DBSMOTE: Density-Based SMOTE; 5.5.7 MWMOTE: Majority Weighted Minority Oversampling TEchnique; Input; Procedure