Advanced support vector machines and kernel methods

Kernel methods (KMs) and support vector machines (SVMs) have become very popular as methods for learning from examples. The basic theory is well understood and applications work successfully in practice. Initially illustrated by their use in classification and regression tasks, recent advanced techniques are presented and key applications are described. Issues of numerical optimization, working set selection, improved generalization, model selection, and parameter tuning are addressed. Application research covering the use of SVMs in text categorization, computer vision, and bioinformatics is discussed. (C) 2003 Published by Elsevier B.V.