摘要
由于信息的不完全性,我们得到的表示事物特征的数据往往是一些区间值.基于区间值的模糊聚类分析方法,可以更大程度地保留信息.首次提出了基于区间值的最大最小法构造相似系数,给出解决区间数多指标信息聚类问题的计算步骤和基于区间值的聚类方法,最后,通过算例说明了所给出的聚类方法在实际中的应用.
The data of expressing characters of things is often interval value, because of incomplete information. Fuzzy clustering analysis based on interval value can preserve in- formation in a more great extent. In this paper, the method of using maximum-minimum method to get similarity coefficient is fist proposed, then the computation steps are given to solve such problems, with a numerical example set, and then the methods of fuzzy clustering analysis based on interval value are given, direct clustering method, maximal tree method, and net-making method. At last, the applications of the new clustering method as proposed are applied to practice.
出处
《数学的实践与认识》
CSCD
北大核心
2010年第3期107-114,共8页
Mathematics in Practice and Theory
基金
河南省科技厅基金(0413031920)
河南省教育厅基金(2008A110021)
关键词
区间值
可能度
最大最小法
编网法
interval value
possibility degree
maximum-minimum method
net-making method
