PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Erratum: SGDQN is Less Careful than Expected
Antoine Bordes, Leon Bottou, Patrick Gallinari, Jonathan Chang and S-Alex Smith
Journal of Machine Learning Research Volume 11, pp. 2229-2240, 2010.

Abstract

The SGD-QN algorithm described in Bordes et al. (2009) contains a subtle flaw that prevents it from reaching its design goals. Yet the flawed SGD-QN algorithm has worked well enough to be a winner of the first Pascal Large Scale Learning Challenge (Sonnenburg et al., 2008). This document clarifies the situation, proposes a corrected algorithm, and evaluates its performance.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:7110
Deposited By:Antoine Bordes
Deposited On:04 March 2011