A Comparison Study of Validity Indices on Swarm-Intelligence-Based Clustering

摘要: Swarm intelligence has emerged as a worthwhile class of clustering methods due to its convenient implementation, parallel capability, ability avoid local minima, and other advantages. In such applications, validity indices usually operate fitness functions evaluate the qualities obtained clusters. However, are data dependent designed address certain types data, selection different may critically affect cluster quality. Here, we compare performances eight well-known widely used indices, namely, Calinski-Harabasz index, CS Davies-Bouldin Dunn index with two generalized versions, I silhouette statistic on both synthetic real sets in framework differential-evolution-particle-swarm-optimization (DEPSO)-based clustering. DEPSO is hybrid evolutionary algorithm stochastic optimization approach (differential evolution) swarm method (particle optimization) that further increases search capability achieves higher flexibility exploring problem space. According experimental results, find stands out most examined. Meanwhile, suggest users reach their conclusions not just based only one but after considering results several achieve reliable structures.