Data Clustering with Cat Swarm Optimization

In this article, a recent metaheuristic method, cat swarm optimization, is introduced to find the proper clustering of data sets. Two clustering approaches based on cat swarm optimization called Cat Swarm Optimization Clustering (CSOC) and K-harmonic means Cat Swarm Optimization Clustering (KCSOC) are proposed. In the proposed methods, seeking mode and tracing mode are adopted to exploit and explore the solution space. In addition, K-Harmonic Means (KHM) operation is designed to refine the population and accelerate the convergence of the clustering algorithm. Experimental results on six real life data sets are given to illustrate the effectiveness of the proposed algorithms.