An Evolutionary Clustering Algorithm Based on Adaptive Fuzzy Weighted Sum Validity Function

摘要

In this paper, we propose a novel objective function called the adaptive Fuzzy Weighted Sum Validity Function (FWSVF), which is a merged weight of the several fuzzy cluster validity functions, including XB, PE, PC and PBMF. The improved validity function is more efficient than others. Furthermore, we present a Mixed Strategy Evolutionary Clustering Algorithm based adaptive validity function (AMSECA), which is merged from Evolutionary Algorithm along with Mixed Strategy and Fuzzy C-means Algorithm. Moreover, in the experiments, we show the effectiveness of AMSECA, AMSECA could find the proper number of clusters automatically as well as appropriate partitions of the data set and avoid local optima.