Adaptive Population of Gaussian Dynamic PSO Clustering Algorithm
自适应种群的高斯动态粒子群聚类算法

The key issue in Clustering is the definition of similarity between samples and the evaluation of pros and cons of clustering effects. PSO algorithm has drawn more attention from the majority of researchers for its preferable impact. This paper gives a new function that measures the effectiveness of the clustering algorithm and analyzes it thoroughly. In addition, from the topology of the PSO, an adaptive population of Gaussian dynamic PSO clustering algorithm is proposed based on the Gaussian dynamic algorithm. The experiment shows the measure function could effectively evaluate the pros and cons of clustering effects, and its corresponding algorithm has good clustering efficiency, better performance in the high-dimensional data.