An efficient clustering approach based on hybridization of PSO, Fuzzy logic and K-harmonic means

Clustering involves formation of groups having similar data objects. The popular clustering techniques k-means and K-harmonic means (KHM) suffered with the problem of local optimum. This paper introduces a new hybrid clustering approach using PSO, Fuzzy logic and KHM. The proposed approach works has two advantages. First it solves the local optimum problem and second is provide better convergence rate as PSOKHM. The performance of proposed approach is evaluated on five benchmark datasets. Experimental results indicate the superiority of the proposed approach.