Keywords

high-order correlations, feature selection, learning theory

Abstract

Many learning algorithms attempt, either explicitly or implicitly, to discover useful high-order features. When considering all possible functions that could be encountered, no particular type of high-order feature should be more useful than any other. However, this paper presents arguments and empirical results that suggest that for the learning problems typically encountered in practice, some high-order features may be more useful than others.

Original Publication Citation

Adam Drake and Dan Ventura, "Comparing High-Order Boolean Features", Proceedings of the Joint Conference on Information Sciences, pp. 428-431, July 25.