An optimization method for characterizing two groups of data

Abstract

Feature selection aims to choose a subset of features, out of the set of candidate features, such that the selected set best represents the whole in a particular aspect. We propose an optimization model for the problem of selecting features in the presence of two groups of data. The objectives include minimizing the number of selected features and maximizing similarities within the same group and differences between different groups. We use the lexicographic method and prove several properties of the problem. We show that obtaining even feasible solutions for the problem can be challenging. Therefore, we propose matheuristic algorithms. We test our algorithms on both randomly generated and real‐world datasets. The computational results indicate that the proposed matheuristics deliver quality solutions in a reasonable amount of time.