关键词: Vague关系, 最小二分割, 最优树, 聚类分析

Abstract: Clustering analysis is a very important problem in the research and development of intelligence system. In order to mitigate the data “distortion” degree caused by the traditional clustering analysis based on equivalence relations, the optimal tree Vague graph clustering method is proposed which introduces the concept of Vague trees and designs the Vague optimum tree by caculating the minimum of two segmentation.  Compared with clustering analysis based on equivalence relations and directive clustering method, the result shows that the method not only effectively reduces the “noise”, but also has the characteristics of continuity and comprehensiveness and more reasonable level of clustering.

Key words: Vague relation, minimum of two segmentation, optimum tree, clustering analysis