colt05.pdf (application/pdf Object)

colt05.pdf (application/pdf Object)

This is the paper we're now discussing. I will post more about it, including comments and questions, as I read it.

Welcome

The ultimate goal of this forum is the following: attempting to understand from a mathematical perspective the results that researchers in the field of machine learning have "discovered" by experiments, and to put these results within a rigorous mathematical framework As a result, we will consider both theoretical works on machine learning and closely related mathematical fields such as functional analysis, and also papers in specific applications of machine learning, for example, to mathematical biology or pattern recognition.

The immediate purpose of the forum is to allow geographically separated researchers to take part in a dicsussion of the following work:

Author: Schölkopf, Bernhard.
Title: Learning with kernels : support vector machines, regularization, optimization, and beyond / Bernhard Schölkopf, Alexander J. Smola.
Published: Cambridge, Mass. : MIT Press, c2002.
Description: xviii, 626 p. : ill. ; 27 cm.

Three researchers, all mathematicians who are not currently experts in machine learning, are meeting each week face-to-face to discuss this text. Several others with a similar background will be taking part remotely. As the group progresses, the members will be encouraged to introduce other works for consideration.