一种用于非球形分布数据的新型聚类算法

A Novel Clustering Algorithm for Data With Non-spherical-shape Distribution

提出了一种新型的模糊遗传聚类算法（ＦＧＣＡ）。该算法不需要预知距离门限ｄ就可对未知类别数目的数据集进行聚类。仿真结果表明，该算法对非球型分布数据同样具有很好的聚类效果。

A novel fuzzy genetic clustering algorithm is presented in this paper.There is no need to know the distance threshold (d) and prior number of clusters while using this algorithm in solving clustering problem. The simulation shows that this algorithm is also effective to non-spherical-shape data.