摘要： 提出一个新的模糊聚类有效性指标。该指标能确定由模糊C-均值算法(FCM)所得模糊划分的最优划分和最优聚类数，结合了模糊聚类的紧致性和分离性信息，用类内加权平方误差和计算紧致性，用类间相似度计算分离性。在3个人造数据集和3个真实数据集上进行对比实验，结果证明该指标的性能优于其他有效性指标。

关键词: 模糊聚类, 有效性指标, 模糊C-均值算法

Abstract: This paper proposes a novel validity index for fuzzy clustering. This index can determine the optimal partition and optimal number of clusters for fuzzy partitions obtained from the Fuzzy C-Means algorithm(FCM). It combines compactness and separation information of fuzzy clustering. The compactness is obtained by computing inter-cluster weighted the square of error. The separation is obtained by computing similarity between fuzzy clustering. Comparison experiment is done in three synthetical datasets and three real datasets, and the results prove that this index is superior effectiveness compared with other validity indexes.

Key words: fuzzy clustering, validity index, Fuzzy C-Means algorithm(FCM)

中图分类号: 

• TP391