Consistency based feature selection

Abstract

Feature selection is an effective technique in dealing with dimensionality reduction for classification task, a main component of data mining. It searches for ein "optimal" subset of features. The search strategies under consideration are one of the three: complete, heuristic, and probabilistic. Existing algorithms adopt various measiires to evaluate the goodness of feature subsets. This work focuses on one measure called consistency. We study its properties in compsirison with other major measures and different ways of using this measure in sejirch of feature subsets. We conduct cin empirical study to examine the pros and cons of these different search methods using consistency. Through this extensive exercise, we ciim to provide a comprehensive view of this measure and its relations with other measures cind a guideline of the use of this meeisure with different search strategies facing a new application.