Gaussian kernel particle swarm optimization clustering algorithm

Abstract

As the K-means algorithm is dependent on the initial clustering center, and the particle swarm optimization (PSO) converges prematurely and is easily trapped in local minima, a Gaussian kernel particle swarm optimization clustering algorithm is proposed in this paper. The algorithm adopts the theory of good point set to initialize population, which makes the initial clustering center more rational. Particle swarm iteration formula was optimized by using Gaussian kernel method, which makes particle swarm algorithm converge rapidly to the global optimal. By testing 23 UCI data sets, the experimental results show that the clustering effect of the proposed algorithm is better than that of the K-means and the traditional particle swarm optimization clustering algorithm.