A new approach based on enhanced PSO with neighborhood search for data clustering

摘要

The well-known K-means algorithm has been successfully applied to many practical clustering problems, but it has some drawbacks such as local optimal convergence and sensitivity to initial points. Particle swarm optimization algorithm (PSO) is one of the swarm intelligent algorithms, it is applied in solving global optimization problems. An integration of enhanced PSO and K-means algorithm is becoming one of the popular strategies for solving clustering problems. In this study, an approach based on PSO and K-means is presented (denoted EPSO), in which PSO is enhanced by neighborhood search strategies. By hybrid with enhanced PSO, it does not only help the algorithm escape from local optima but also overcomes the shortcoming of the slow convergence speed of the PSO algorithm. Experimental results on eight benchmark data sets show that the proposed approach outperforms some other data clustering algorithms, and has an acceptable efficiency and robustness.