Harmony search algorithm for simultaneous clustering and feature selection

Feature selection is an important process in data analysis for information-preserving data reduction. Clustering is inherently a difficult task and is made even more difficult when the selection of relevant features is also an issue. In this paper, we propose an approach for clustering and feature selection simultaneously using a harmony search algorithm. Our approach makes feature selection an integral part of the global clustering search procedure and attempts to overcome the problem of identifying less promising locally optimal solutions in both feature selection and clustering, without making any a prior assumption about the number of clusters. Within ourmethod, a variable composite representation is devised toencode both feature selection and cluster centers with different numbers of clusters. Furthermore, local search operations are used to improve feature selection and cluster centers encoded in the harmonics.