Introducing Hybrid model for Data Clustering using K-Harmonic Means and Gravitational Search Algorithms

Clustering is a process of extracting reliable, unique, effective and comprehensible patterns from database. Various clustering methods are proposed to accomplish exactness and accuracy of clusters. K-Means is well known clustering algorithm but it easily converge to local optima. To overcome this drawback, an improved algorithm called K-Harmonic Mean (KHM) was proposed, which is independent of cluster center initialization. This article presents study of hybridization KHM with other clustering algorithms. In order to improve the clustering accuracy the authors proposed new hybrid KHM model.