Improving Statistical Measures of Feature Subsets by Conventional and Evolutionary Approaches

Abstract

Abstract. In this paper we compare recently developed and highly effective sequential feature selection algorithms with approaches based on evolutionary algorithms enabling parallel feature subset selection. We introduce the oscillating search method, employ permutation encoding offering some advantages over the more traditional bitmap encoding for the evolutionary search, and compare these algorithms to the often studied and well-performing sequential forward floating search. For the empirical analysis of these algorithms we utilize three well-known benchmark problems, and assess the quality of feature subsets by means of the statistical Bhattacharyya distance measure.

Identifying the most Informative Variables for Decision-Making Problems - a Survey of Recent Approaches and Accompanying Problems

“…It may also help to find solutions in a significantly shorter time, although this is not guaranteed. The optimization power of purely randomized procedures like genetic algorithms [10], [19], [39] has been found slightly inferior to sequential methods. Extending sequential methods to include limited randomization may be a good compromise, as is the case with repeatedly randomly initialized oscillating search [33].…”

A Framework for Classifier Fusion: Is It Still Needed?

“…It has been also shown that feature selection can be performed as part of the process of data modelling using gaussian mixtures [10,13]. The effectiveness of evolutionary optimisation approaches in feature selection has been demonstrated in [18,17,19]. The use of fused classifier error as a criterion for feature selection has been suggested in [20,21].…”