聚类分析是数据挖掘的一个重要研究课题,模糊聚类是聚类分析的一个有效手段。本文在分析AFS方法和FCM算法的基础上,设计了一个基于AFS拓扑和FCM的模糊聚类算法,进行iris数据的聚类实验证明它聚类结果优于传统的FCM聚类算法,具有很好的推广性和实际应用价值。
Cluster analysis is an important research topic in data mining (DM), fuzzy cluster analysis is an effective method of cluster analysis. This paper proposes a novel clustering algorithm after analyzing the AFS theory and FCM algorithm. The experimenttal results show that it is not only better than tradition FCM algorithm but also has good practical value and can be used into other field easily.