混合类型数据的聚类算法综述

摘要

Among all the algorithms of data mining, clustering algorithm is especially important. In practical applications, most data appear in mixed data with both numerical and categorical attributes. Therefore, the clustering analysis of mixed data attracts more and more attention. The author introduces the definition of clustering, the basic principles of K-prototype clustering algorithm, some clustering algorithms of mixed data by researchers in recent years, and gives a brief introduction to its specific applications.%在所有数据挖掘的算法中,聚类算法尤为重要.在实际应用中,数据大多是混合型的数据,既包含数值型,又包含分类型数据.因此,对混合数据进行聚类分析越来越重要.笔者首先介绍了聚类的定义,然后详细阐述了K-prototype聚类算法的基本原理,最后详细介绍了近年来各学者对混合类型数据的聚类算法的研究现状和其具体应用.