摘要： 介绍了一种基于密度和网格的聚类分析算法--蚁群爬山法(ACH)，这种算法能自动获得簇数k的值和任意形状的簇的划分，并具有较好的并行性。通过对网格大小的控制可获得不同层次的聚类结果。

关键词: 蚁群爬山法, 聚类, 算法

Abstract: The paper introduces a kind of clustering algorithm based on density and grids--ants climb hill algorithm.It can get value of k and cluster with different shapes together. It has good parallel ability. The size of the grids decides precision of the clustering.

中图分类号: 

• TP393.02