基于数学形态学的聚类集成算法

Clustering Ensemble Algorithm Based on Mathematical Morphology

提出了基于数学形态学的聚类集成算法CEOMM。它利用不同的结构元素的探针作用,对不同的结构元素探测出来的簇核心图进行集成,在集成所得到的簇核心基础上聚类。实验结果表明,算法CEOMM对有复杂类形状的数据集进行聚类时,效果比传统聚类算法更好,且能确定聚类数。而且由于采用了不同的结构元素进行探测,对于由不同形状的类构成的数据集其聚类效果很理想。

In this paper, a clustering ensemble algorithm named CEOMM was proposed, which combines multiple clustering cores explored by different structure elements to get a desirable and correct clustering core of a data set. And then CEOMM gets the clustering of the data set based on the ensemble clustering core. Experimental results demonstrate CEOMM can cluster data with complex cluster shapes better than the classical clustering algorithms, and it can also find an optimal number of clusters. Moreover, CEOMM can discover overlapping clusters with different arbitrary shapes, because it uses different structure elements.