摘要
基于硬聚类算法的几种有效性指标,即Hubert统计量、Davies-Bouldin指标、Dunn's指标以及Dunn's指标的推广,提出了相应的适用于模糊聚类算法的有效性指标.实验证明,这些改进的有效性函数对模糊C-均值算法而言同样有效.
The paper presents some fuzzy validity criteria based on well-known hard clustering validation methods: Hubert statistic, Davies- Bouldin index, Dunn's index, and several generalizations of an index due to Dunn. The numerical experiments show that these modified validity functions are efficient for fuzzy clustering.
出处
《江南大学学报(自然科学版)》
CAS
2007年第6期878-882,共5页
Joural of Jiangnan University (Natural Science Edition)