Abstract:

The results of the HCM clustering algorithm are often local optimal solutions. There often exists insignificant clustering in the result of the HCM cl ustering algorithm when traditional definitions are adopted in the operations between fuzzy sets. Our research indicates that some local optimal solutio ns can be avoided by using genetic algorithms in the HCM clustering algorithm. Therefore this article applies genetic algorithms to the HCM clustering a  lgorithm and designs the corresponding algorithms. Considering the efficiency and overhead, this paper modifies this algorithm. Finally, this paper puts  forward a new algorithm named CHCM clustering. And then it compares the performance of CHCM with HCM using the test data. Experimental results show that the performance of CHCM is far better than that of HCM.

Key words: HCM clustering algorithm;CHCM clustering algorithm;GA