Data clustering using higher order statistics

Abstract

Traditional k-means algorithms for data clustering are based on the assumption that the underlying distribution of the data is Gaussian. In this paper, we propose a new clustering algorithm that makes use of higher order statistics for improved data clustering when the distribution of the data is non-Gaussian. The algorithm uses an HOS-based decision measure which is derived from a series expansion of the multivariate probability density function in terms of the multivariate Gaussian and the Hermite polynomials. Experimental results are presented on the performance of the proposed algorithm.