基于商空间和信息粒度的Fuzzy聚类分析

Fuzzy Clustering Analysis Based on Quotient Spaceand Information Granularity

传统聚类分析主要从相似性度量的不同设计和判别函数的选择等方面进行研究。本文从商空间理论和信息粒度原理角度出发,引入分层递阶结构,论述聚类分析的本质以及Fuzzy聚类分析这种软件统计方法。

Traditional clustering analyses mainly study the designing of similarity measurement and the selec-(ting) of discriminant function. In this paper, the authors define the hierarchy based on quotient space and information granularity; discuss the nature of clustering analysis and fuzzy clustering analysis.