Vague相似max-t_Δ传递性模糊聚类  

Vague cluster analysis based on vague smilar and max-t_Δ transitivity

作　　者：陈均明[1] 李红霞[2] 

机构地区：[1]重庆工商大学数学与统计学院,重庆400067 [2]重庆工商大学管理学院,重庆400067

出　　处：《计算机工程与应用》2010年第35期159-162,共4页Computer Engineering and Applications

摘　　要：提出同时考虑真隶属度之差、假隶属度之差、得分值之差和清晰度之和的Vague值相似度量方法。将Vague聚类转化为经典模糊集聚类:先计算Vague对象两两之间的相似度,得到经典模糊相似矩阵,然后运用max-tΔ传递性进行聚类。这种聚类法计算更简单,丢失信息更少,聚类更准确。A new similar method of Vague sets is proposed which based on the difference of true membership degree and of false membership degree and of scores and on the sum of definition.Then clustering in vague sets are transfered into in fuzzy setst,hat isf,irst calculate the similar degree between two Vague objects and get a classical fuzzy similar matrixt,hen utilize max-tΔ transitivity for fuzzy clustering.This clusting method makes that calculating is more simplet,he loss informa-tion is more less,and the clusting is more exact.

关 键 词：相似度 max-tΔ传递性 模糊等价关系 聚类 丢失信息 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]