[.ca] Advances in Kernel Methods: Support Vector Learning (ISBN 0262194163)

a summary of research on support vector machines:
This is a collection of papers presented at a NIPS workshop held in 1997. So it provides a good entry point for access to forefronts of this rapidly developing field. Many leading researchers have contributed to this volume including V. vapnik who wrote a very succinct and readable survey. The introduction (Chapter 1) is also very useful. Though all chapters are written by leading experts in their areas and are enjoy to read. Personally I like particularly Part II on implementation in large data sets. G. Wahba provides some background on RKHS theory and a statistical perspective from GACV, for which she is mainly responsible for its popularity in statistics. I recommend this book for researchers and practitioners who may want more details and update recent developments.