arXiv:cond-mat/0010423 Abstract

arXiv:cond-mat/0010423Abstract References Reviews Resources

Hierarchical learning in polynomial Support Vector Machines

Published 2000-10-26Version 1

We study the typical properties of polynomial Support Vector Machines within a Statistical Mechanics approach that allows us to analyze the effect of different normalizations of the features. If the normalization is adecuately chosen, there is a hierarchical learning of features of increasing order as a function of the training set size.

Comments: 22 pages, 7 figures, submitted to Machine Learning

Categories: cond-mat.dis-nn

Keywords: polynomial support vector machines, hierarchical learning, statistical mechanics approach, normalization, typical properties

Related articles: Most relevant | Search more

arXiv:cond-mat/9808039 (Published 1998-08-04, updated 1998-12-14)

How to Implement A Priori Information: A Statistical Mechanics Approach

Joerg C. Lemm

arXiv:cond-mat/0703191 (Published 2007-03-07, updated 2007-03-08)

A statistical mechanics approach for scale-free networks and finite-scale networks

Ginestra Bianconi

arXiv:cond-mat/0307236 (Published 2003-07-10)

Statistical mechanics of the vertex-cover problem

Alexander K. Hartmann, Martin Weigt

arXiv Analytics