期刊论文

Gravity-based clustering approach

中文

计算机应用

1001-9081

2005

25

2

286-288,300

国家自然科学基金

本校为第一机构

将万有引力的思想引入聚类分析中,提出了一种基于引力的聚类方法GCA(Grayity-based Clustering Approach),同时给出了一种计算聚类阈值的简单而有效的方法.GCA关于数据库的大小和属性个数具有近似线性时间复杂度,这使得聚类方法GCA具有好的扩展性.实验结果表明GCA可产生高质量的聚类结果.

The idea of the law of gravity was introduced to clustering analysis, and a gravity-based clustering approach, named GCA, was presented in this paper. At the same time, a simple method to calculate cluster threshold was put forward. The clustering approach GCA had the nearly linear time complexity with the size of dataset and the number of attributes, which resulted in good expandability. The experimental results ...