Automatic clustering approach based on particle swarm optimization for data with arbitrary shaped clusters
Geng-Bin Chen, An Song, Chun-Ju Zhang, Xiaofang Liu, Wei–Neng Chen, Zhi‐Hui Zhan, Jinghui Zhong, Jun Zhang, Xiao-Min Hu · 2016
Recently, partitional clustering approaches based on Evolutionary Algorithms (EAs) have shown promising in solving the data clustering problems. However, with the nearest prototype (NP) rule as the method for decoding, most of them are only suitable for clustering datasets with convex (e.g. hyperspherical) clusters. In this paper, we propose an automatic clustering approach using particle swarm optimization (PSO). A new encoding scheme with a novel decoding method, named the nearest multiple prototypes (NMP) rule, is applied to the PSO-based clustering algorithm to automatically determine an appropriate number of clusters in the procedure of clustering and partition datasets with arbitrary shaped clusters. The algorithm is experimentally validated on both synthetic and real datasets. The results show that the proposed PSO-based approach is very competitive when comparing with two popular clustering algorithms.