一种基于区间数多指标信息的聚类方法

A Clustering Method Based on Interval Numbers for Multi-index Information

针对一类特征指标值及指标权重均为区间数的多指标信息聚类问题,给出了一种最大树聚类分析方法·首先对区间数多指标信息聚类问题进行了描述;然后依据传统的基于数值信息的最大树模糊聚类分析方法的基本思路,给出了解决区间数多指标信息聚类问题的计算步骤·最后,通过给出一个算例说明了所给出的聚类方法·

With respect to the problems of which both the characteristic values and weights of indices are uncertain interval numbers, a new maximal tree clustering analysis method is proposed. After a discussion on the problems of multi-index information clustering analysis in terms of interval numbers, a basic concept is formed according to the traditional maximal tree clustering method based on numerical information. Then the computation steps are given to solve such problems, with a numerical example set out to illustrate the applications of the new clustering method as proposed.