Cost-Sensitive Machine Learning

In machine learning applications, practitioners must take into account the cost associated with the algorithm. These costs include:


Cost of acquiring training data
Cost of data annotation/labeling and cleaning
Computational cost for model fitting, validation, and testing
Cost of collecting features/attributes for test data
Cost of user feedback collection
Cost of incorrect prediction/classification

Cost-Sensitive Machine Learning is one of the first books to provide an overview of the current research efforts and problems in this area. It discusses real-world applications that incorporate the cost of learning into the modeling process.

The first part of the book presents the theoretical underpinnings of cost-sensitive machine learning. It describes well-established machine learning approaches for reducing data acquisition costs during training as well as approaches for reducing costs when systems must make predictions for new samples. The second part covers real-world applications that effectively trade off different types of costs. These applications not only use traditional machine learning approaches, but they also incorporate cutting-edge research that advances beyond the constraining assumptions by analyzing the application needs from first principles.

Spurring further research on several open problems, this volume highlights the often implicit assumptions in machine learning techniques that were not fully understood in the past. The book also illustrates the commercial importance of cost-sensitive machine learning through its coverage of the rapid application developments made by leading companies and academic research labs.

Balaji Krishnapuram is a senior R&D manager at Siemens Medical Solutions. He earned a Ph.D. in electrical and computer engineering from Duke University. His research interests include statistical data mining and information retrieval.
Shipeng Yu is a senior staff scientist at Siemens Medical Solutions. He earned a Ph.D. in computer science from the University of Munich. His research interests include statistical machine learning, data mining, Bayesian analysis, information retrieval and extraction, healthcare analytics, and personalized medicine.
R. Bharat Rao is senior director and head of Knowledge Solutions at Siemens Medical Solutions, where was recognized as one of its Inventors of the Year in 2005. He also received the 2011 ACM SIGKDD Lifetime Service Award for pioneering applications of data mining for healthcare. He earned a Ph.D. in electrical and computer engineering from the University of Illinois at Urbana-Champaign. His research interests include machine learning, healthcare analytics, mining large data, and personalized medicine.

THEORECTICAL UNDERPINNINGS OF COST-SENSTIVE MACHINE LEARNING Algorithms for Active Learning, Burr SettlesQuery Strategy FrameworksA Unified View Summary and Outlook

Semi-Supervised Learning: Some Recent Advances, Xueyuan Zhou, Ankan Saha, and Vikas SindhwaniSemi-Supervised Prediction for Structured OutputsTheoretical AnalysisNew Directions

Transfer Learning, Multi-Task Learning, and Cost-Sensitive Learning, Bin Cao, Yu Zhang, and Qiang YangNotations Transfer Learning ModelsMulti-Task Learning ModelsConclusion and Future Work

Cost-Sensitive Cascades, Vikas C. RaykarFeatures Incur a CostCascade of Classifiers Successful Applications of Cascaded ArchitecturesTraining a Cascade of ClassifiersTradeoff between Accuracy and Cost Conclusions and Future Work

Selective Data Acquisition for Machine Learning, Josh Attenberg, Prem Melville, Foster Provost, and Maytal Saar-TsechanskyOverarching Principles for Selective Data AcquisitionActive Feature-Value Acquisition Labeling Features versus ExamplesDealing with Noisy AcquisitionPrediction Time Information AcquisitionAlternative Acquisition SettingsConclusion

COST-SENSITIVE MACHINE LEARNING APPLICATIONS Minimizing Annotation Costs in Visual Category Learning, Sudheendra Vijayanarasimhan and Kristen GraumanReducing the Level of SupervisionReducing the Amount of SupervisionReducing the Effort Required in SupervisionCost-Sensitive Multi-Level Active LearningConclusion

Reliability and Redundancy: Reducing Error Cost in Medical Imaging, X.S. Zhou, Y. Zhan, Z. Peng, M. Dewan, B. Jian, A. Krishnan, M. Harder, R. Schwarz, L. Lauer, H. Meyer, S. Grosskopf, U. Feuerlein, H. Ditt, and M. ScheueringA Measure of ReliabilityReliability of Pattern Localization: Asymmetric Cost for FPs and FNs Implications and Learning Strategy for Medical Imaging ApplicationsRelated Work and Discussions

Cost-Sensitive Learning in Computational Advertising, Deepak AgarwalPerformance Advertising: Sponsored Search and Contextual MatchingDisplay AdvertisingDiscussion

Cost-Sensitive Machine Learning for Information Retrieval, Martin Szummer and Filip RadlinskiUtility in Information RetrievalLearning to RankReducing Labeling CostMultiple Utilities Conclusion

Index

A Bibliography appears at the end of each chapter.