摘要
粗糙C均值算法中3个参数wl,wu,ε的选择是算法应用的关键问题。针对粗糙C均值算法中反映类间叠加程度的参数ε的设定,提出一种动态自适应调整阈值ε的粗糙C均值算法,该算法根据"类-类"间距离与"对象-类"间距离,对每一个待聚类对象动态设定阈值ε。两组人工数据和图像数据的实验表明,该算法具有较好的适应性和聚类效果。
Selection of parameters w_l,w_u,ε plays an important role in rough C-Means algorithm.In this paper,a dynamicthreshold rough C-Means algorithm was proposed to self-adaptive adjusting threshold ε that reflects the superposition between classes.This algorithm computes a threshold for every object on the basis of class interval and the distance between class and object.The better effect can be testified by two synthetic data and image data experiments.
出处
《计算机科学》
CSCD
北大核心
2011年第3期218-221,242,共5页
Computer Science
基金
国家自然科学基金(60872152)资助
关键词
C均值聚类
粗糙集
粗糙C均值聚类
C-Means clustering
Rough sets
Rough C-Means clustering