摘要
聚类算法是数据挖掘中的一个重要研究领域,是一种数据划分或分组处理的重要手段和方法.目前其研究已深入到数据库、数据挖掘、统计等领域并取得了很大的成绩,但是由于聚类算法的多样性,使其在很多行业应用中有着不同的应用效果,基于此,本文通过聚类算法三种指标的比较,给出了一种聚类方法应用效果评估的方法.该方法结合电信的案例应用与K-Means、SOM、BIRCH等聚类方法结果的分析,最后得出K-Means方法在电信客户细分中的应用优越性.
Clustering algorithm is an important area of research in a data mining, and is an important means and methods of data partitioning or packet processing. At present this research have been deep into the database, data mining, statistics and other areas and it has made great achievements, but the diversity of clustering algorithm make it have different application effect in many industry＇s ap- plications. Based on this, this paper presents a method of clustering method application effect assessment by comparison of three indicators of clustering algorithm. This method combines the case applications of telecom with the resuh analysis of clustering methods such as K-Means, SOM, BIRCH ,rod so on. Finally, it obtains the superiority of K-Means approach applied to the telecnmmunication＇s customer segmentation.
出处
《邵阳学院学报（自然科学版）》
2009年第4期69-73,共5页
Journal of Shaoyang University：Natural Science Edition
关键词
聚类算法
区分度
效果评估
clustering algorithm
differentiation
effect assessment
