摘要
针对现有数据流聚类算法的不足,提出了改进模糊聚类算法,给数据对象的隶属度加上一个权值,以及在算法中用有效性函数来确定聚类数目C.理论分析与实验结果表明,在数据流环境下所提出的改进模糊聚类算法比传统算法有更好的聚类效果,更快的聚类速度.
In order to overcome the shortrage of data stream clustering, we present a improved clustering algorithm. The subjection value of the algorithm is modified by adding weighted value and the choice for parameter of number of clusters based on cluster validity function. Experiments show that the FCM has a better cluster result and has faster clustering rate.
出处
《河南科学》
2012年第9期1243-1245,共3页
Henan Science
关键词
数据流
模糊
聚类
data stream： fuzzy
clustering
