Multiobjective optimization based subspace clustering using evolvable genome structure

Dipanjyoti Paul, Sriparna Saha, Jimson Mathew

Abstract: Subspace clustering techniques become popular in identifying local patterns from high dimensional data. In this paper, we present a multiobjective optimization based evolutionary algorithm in order to tackle the subspace clustering problem. Previous state-of-the-art algorithms on subspace clustering optimize implicitly or explicitly a single cluster quality measure. The proposed approach optimizes two cluster quality measures namely PBM-index and XB-index simultaneously. The developed algorithm is applied to seven standard real-life datasets for identifying different subspace clusters. Experimentation reveals that the proposed algorithm is able to take advantages of its evolvable genomic structure and multiobjective based framework and it can be applied to any data set.