Support Vector Machines: Optimization Based Theory, Algorithms, and Extensions

Providing a comprehensive introduction to support vector machines (SVMs) from an optimization and operations research perspective, this book highlights the substantial and growing interest in machine learning within the optimization community. It describes the classification and regression problems that make up the two main components of SVMs and constructs SVMs for solving multi-class classification problems and handling multi-instance problems. Topics covered include probability for SVMs, twin-SVMs, SVMs with privileged information, multi-label SVMs, AVC maximizing SVMs, and kernel PCA.