PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Spectral Algorithms for Supervised Learning
Laura Lo Gerfo, Lorenzo Rosasco, Francesca Odone, Ernesto De vito and Alessandro Verri
neural computation Volume to appear, 2007.

Abstract

We discuss how a large class of regularization methods, collectively known as spectral regularization and originally designed for solving ill-posed inverse problems, gives rise to regularized learning algorithms. All these algorithms are consistent kernel methods which can be easily implemented. The intuition behind their derivation is that the same principle allowing to numerically stabilize a matrix inversion problem is crucial to avoid over-fitting. The various methods have a common derivation, but different computational and theoretical properties. We describe examples of such algorithms, analyzing their classification performance on several datasets and discussing their applicability to real world problems.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:3945
Deposited By:Lorenzo Rosasco
Deposited On:25 February 2008