面向层次类型变量的相异度量及其聚类分析

Dissimilarity Metric and Clustering for Hierarchy Variable

本文在分析传统类型变量相异度量的基础上,定义了"层次类型"的概念,提出了层次类型变量的相异度量计算方法。引入层次类型变量,并结合传统类型变量,设计了具有包括层次类型在内的混合数据类型描述的对象之间的相异度量方法,并基于此实现了此类对象的聚类分析。

Based on the analysis for the dissimilarity metric of traditional type variables, the hierarchy type is defined, and the dissimilarity metric for the type is proposed in the paper. Moreover, the dissimilarity metric of the hybrid type including traditional types and the hierarchy type is designed, and a clustering algorithm based on the metric is implemented.