摘要
针对离群点数据的发现问题,提出一种改进的离群核模糊聚类算法,利用先验知识,选择聚类目标函数,并将聚类中心作为初始化参数,有效提高算法的收敛速度,减少其整体运行时间,仿真实验结果表明,该算法是有效的。
Aiming at the discovery problem in outlier data, an improved fuzzy clustering algorithm for outlier kernel is proposed, which uses priori knowledge to select clustering objective function and makes clustering center as initial parameter. The convergence rate of the algorithm is promoted and the whole running time is reduced. Simulation experimental results show this algorithm is effective.
出处
《计算机工程》
CAS
CSCD
北大核心
2009年第11期190-192,共3页
Computer Engineering
关键词
离群
模糊聚类
核函数
outlier
fuzzy clustering
kernel function