Imbalanced Learning: Foundations, Algorithms, and Applications

The first book of its kind to review the current status and future direction of the exciting new branch of machine learning/data mining called imbalanced learning Imbalanced learning focuses on how an intelligent system can learn when it is provided with imbalanced data. Solving imbalanced learning problems is critical in numerous data-intensive networked systems, including surveillance, security, Internet, finance, biomedical, defense, and more. Due to the inherent complex characteristics of imbalanced data sets, learning from such data requires new understandings, principles, algorithms, and tools to transform vast amounts of raw data efficiently into information and knowledge representation. The first comprehensive look at this new branch of machine learning, this book offers a critical review of the problem of imbalanced learning, covering the state of the art in techniques, principles, and real-world applications. Featuring contributions from experts in both academia and industry, Imbalanced Learning: Foundations, Algorithms, and Applications provides chapter coverage on: Foundations of Imbalanced Learning Imbalanced Datasets: From Sampling to Classifiers Ensemble Methods for Class Imbalance Learning Class Imbalance Learning Methods for Support Vector Machines Class Imbalance and Active Learning Nonstationary Stream Data Learning with Imbalanced Class Distribution Assessment Metrics for Imbalanced Learning Imbalanced Learning: Foundations, Algorithms, and Applications will help scientists and engineers learn how to tackle the problem of learning from imbalanced datasets, and gain insight into current developments in the field as well as future research directions.

Preface ix Contributors xi 1 Introduction 1 Haibo He 1.1 Problem Formulation, 1 1.2 State–of–the–Art Research, 3 1.3 Looking Ahead: Challenges and Opportunities, 6 1.4 Acknowledgments, 7 References, 8 2 Foundations of Imbalanced Learning 13 Gary M. Weiss 2.1 Introduction, 14 2.2 Background, 14 2.3 Foundational Issues, 19 2.4 Methods for Addressing Imbalanced Data, 26 2.5 Mapping Foundational Issues to Solutions, 35 2.6 Misconceptions About Sampling Methods, 36 2.7 Recommendations and Guidelines, 38 References, 38 3 Imbalanced Datasets: From Sampling to Classifiers 43 T. Ryan Hoens and Nitesh V. Chawla 3.1 Introduction, 43 3.2 Sampling Methods, 44 3.3 Skew–Insensitive Classifiers for Class Imbalance, 49 3.4 Evaluation Metrics, 52 3.5 Discussion, 56 References, 57 4 Ensemble Methods for Class Imbalance Learning 61 Xu–Ying Liu and Zhi–Hua Zhou 4.1 Introduction, 61 4.2 Ensemble Methods, 62 4.3 Ensemble Methods for Class Imbalance Learning, 66 4.4 Empirical Study, 73 4.5 Concluding Remarks, 79 References, 80 5 Class Imbalance Learning Methods for Support Vector Machines 83 Rukshan Batuwita and Vasile Palade 5.1 Introduction, 83 5.2 Introduction to Support Vector Machines, 84 5.3 SVMs and Class Imbalance, 86 5.4 External Imbalance Learning Methods for SVMs: Data Preprocessing Methods, 87 5.5 Internal Imbalance Learning Methods for SVMs: Algorithmic Methods, 88 5.6 Summary, 96 References, 96 6 Class Imbalance and Active Learning 101 Josh Attenberg and S¸eyda Ertekin 6.1 Introduction, 102 6.2 Active Learning for Imbalanced Problems, 103 6.3 Active Learning for Imbalanced Data Classification, 110 6.4 Adaptive Resampling with Active Learning, 122 6.5 Difficulties with Extreme Class Imbalance, 129 6.6 Dealing with Disjunctive Classes, 130 6.7 Starting Cold, 132 6.8 Alternatives to Active Learning for Imbalanced Problems, 133 6.9 Conclusion, 144 References, 145 7 Nonstationary Stream Data Learning with Imbalanced Class Distribution 151 Sheng Chen and Haibo He 7.1 Introduction, 152 7.2 Preliminaries, 154 7.3 Algorithms, 157 7.4 Simulation, 167 7.5 Conclusion, 182 7.6 Acknowledgments, 183 References, 184 8 Assessment Metrics for Imbalanced Learning 187 Nathalie Japkowicz 8.1 Introduction, 187 8.2 A Review of Evaluation Metric Families and their Applicability to the Class Imbalance Problem, 189 8.3 Threshold Metrics: Multiple– Versus Single–Class Focus, 190 8.4 Ranking Methods and Metrics: Taking Uncertainty into Consideration, 196 8.5 Conclusion, 204 8.6 Acknowledgments, 205 References, 205 Index 207

HAIBO HE, PhD, is an Associate Professor in the Department of Electrical, Computer, and Biomedical Engineering at the University of Rhode Island. He received the National Science Foundation (NSF) CAREER Award and Providence Business News (PBN) Rising Star Innovator Award. YUNQIAN MA PhD, is a senior principal research scientist of Honeywell Labs at Honeywell Inter-national, Inc. He received the International Neural Network Society (INNS) Young Investigator Award.

“This book certainly qualifies as a reference for graduate studies in machine learning. Research students are sure to find it highly valuable and a prized possession, especially taking into account the wealth of supporting literature that the authors have brought to the fore.”  ( Computing Reviews , 27 March 2014)