一种自动选取参数的聚类算法

A Clustering Algorithm of Selecting All Parameters Automatically

摘要: 本文提出"K-MEANS"方法的一种聚类算法:ASP.这个算法脱胎于ISODATA,改进与修改了CLASS算法。它可以根据具体的数据自动选取参数,文中介绍了用ASP算法计算的几个实例。

Abstract: This paper presents a clustering algorithm of K-MEANS method: ASP algorithm. Based on the ISODATA the algorithm improves the CLASS algorithm. All parameters can be selected automatically according to the actual data. Some practical examples of using ASP algorithm are introduced in the paper.