摘要
针对不规则形状分布的数据,提出了一种新型模糊聚类算法.该方法结合了近邻函数准则分类算法,对模糊C均值聚类算法进行了拓展.仿真实验表明对球形分布的数据和非球形分布的数据,这种新算法的聚类性能优于模糊均值聚类算法.
A novel fuzzy clustering algorithm is designed for data whose underlying distribution shapes are irregular. This paper is concerned with the association of the fuzzy C-means algorithm and the neighbor function classing algorithm.The simulated results show that this novel algorithm is superior to the normal fuzzy C-means algorithm, for both the spherical-shape distribution data set and the non- spherical-shapedistribution data set.
出处
《空军雷达学院学报》
2002年第2期32-34,共3页
Journal of Air Force Radar Academy
关键词
模糊C均值聚类算法
近邻函数
正向连接损失
fuzzy C-mean algorithm (FCA)
neighbor function
forward-connecting loss
