A Novel and Efficient Rough Set Based Clustering Technique for Gene Expression Data

Abstract

Gene expressions with similar patterns are clustered, which help us to understand the functions of unknown and abnormal patterns of genes in future. The major task of gene expression data clustering is to identify groups of co-expressed genes. In this regard a new gene expression clustering method, termed as A Novel and Efficient Rough Set Based Clustering Technique for Gene Expression Data (NRSBCGE), is proposed based on the Rough set theory. This method is designed intelligently as it itself detects the optimum number of clusters. The proposed clustering method provides an efficient way of finding the unique gene expression patterns. The method was experimented with two publicly available cancer datasets and the results were compared with two existing methods of clustering. The effectiveness of the proposed method, along with a comparison with existing Rough set based gene selection and clustering algorithms, is demonstrated based on the silhouette index, which provides better result than the previously proposed methods.