Scalable Kernel Systems

V. Tresp and A. Schwaighofer

Abstract:

Kernel-based systems are currently very popular approaches to supervised learning. Unfortunately, the computational load for training kernel-based systems increases drastically with the number of training data points. Recently, a number of approximate methods for scaling kernel-based systems to large data sets have been introduced. In this paper we investigate the relationship between three of those approaches and compare their performances experimentally.

Reference: V. Tresp and A. Schwaighofer. Scalable kernel systems. In G. Dorffner, H. Bischof, and K. Hornik, editors, Artificial Neural Networks - ICANN 2001, Lecture Notes in Computer Science2130, pages 285-291. Springer Verlag, 2001.