Comparison of Clustering Methods with Generating Mixture Component Data

摘要：

Clustering is an important research topic and core technology in Data Mining.Clustering algorithms have been researched deeply.NOW.there are lots of different clustering algorithms,they are used in special fields and users.In order to use these algorithms better,some researchers have proposed some standards to evaluate the Clustering algorithms.This paper aims to evaluate Clustering algorithms fouln another aspect-using the Renerating mixture component data sets which have overlapping phenomenon to compare Clustering algorithms’property.Based on the concept of overlap rate,we can Renerate data sets with different geometrical character.Then we use the data sets to evaluate Clustering algorithms to find the applicability of clustering algorithms.

关键词： Data mining Clustering Algorithm Model Mixture Overlap Rate

作者: Mei SUN Lingjun KONG Haojun SUN

会议类型: 国际会议

会议名称: The International Conference Information Computing and Automation(2007国际信息计算与自动化会议)

会议地点: 成都

会议语种:英文

页码: 639-642

在线出版日期: 2007-12-19（万方平台首次上网日期，不代表论文的发表时间）