一种面向非规则非致密空间分布数据的聚类方法

Clustering Method for Irregular and Uncompact Data

针对目前很少关注非致密非规则数据聚类的情况,利用蚁群算法具有的组合优化方面的优势,引入近邻函数准则,提出了基于蚁群算法和近邻函数准则的聚类算法,来求解非规则非致密数据聚类问题。实验表明,对于非规则非致密分布数据的聚类问题,该聚类算法可根据连接关系合理地进行聚类,相比K均值算法等其他采用样本距离作为分类指标的聚类方法,可有效降低错聚率,一定程度上较好地解决了这类问题。

Taking advantage of ant colony algorithm＇s superiority on combinatorial optimization problems, a new clustering algorithm based on ant colony algorithm and neighbor function criterion was presented for uncompact and irregu lar data. Result shows that this new clustering algorithm can obtain better clustering results than K-means clustering algorithm and can commendably solve the problem about the uncompact and irregular distribution to a certain extent.