一种新的加权模糊C中心聚类算法
首发时间:2009-06-29
摘要:对于一些局部分布稀疏不均、聚类区域的形状及大小很不规整的数据点集,多数聚类算法不能很好地探测出其聚类分布。在借鉴了两个加权FCM聚类算法的构造及推导过程的基础上,提出了一种新的加权模糊C中心聚类算法(算法1)。接着对该聚类算法进行了一些讨论,给出其时间复杂度及收敛性分析。通过German数据集的几种聚类算法的对照实验结果及评估相异性度量的比较实验结果,验证了该聚类算法有时能取得比WFCM更好的聚类精度,从而说明这个新型加权聚类算法具有一定的有效性。最后给出了几点研究展望,为下一步的研究指明了方向。
For information in English, please click here
An New Weighted Fuzzy C Centers Clustering Algorithm
Abstract:Many clustering algorithms can not explore clustering distributing of some data-point sets with erose shape and size, which are sparse and asymmetrical local distributing. In order to solve this problem, it gives a new weighted fuzzy c centers clustering algorithm(WFCC), described by algorithm 1, based on constructing and deducing of two weighted clustering algorithms. After some discuss for WFCC, it lists its time complexity and astringency analysis. It validates the new clustering algorithm can often get a better clustering quality by an experiment using German data sets. At last, it indicates several valuable research expectations.
Keywords: Weighted Clustering Ordered Attributes Sorted Attributes Hybrid Attributes
基金:
论文图表:
引用
No.3349222051124626****
同行评议
共计0人参与
勘误表
一种新的加权模糊C中心聚类算法
评论
全部评论0/1000