广西工学院学报2012，Vol.23Issue(1)：24-27,33,5.
基于网格的二次K-means聚类算法
Two times K-means algorithm based on grid
摘要
Abstract
Classical K-means is a popular clustering algorithm,but it＇s sensitive to initial mean points,and is mostly influenced by noisy and abnormal data.So the paper provides a two times K-means algorithm based on grid.Firstly,the algorithm divides the space to many equal grids,and then gets dense grid.The algorithm deals with the points in dense grid to firstly clustering.In secondly clustering,the algorithm uses the mean points that are results of firstly clustering as initial mean points of second times.So it can remove the influence of noisy and abnormal data,and keep the completeness of information.Experiments prove the algorithm is effective.关键词
数据挖掘,聚类/K-均值算法/网格Key words
data mining/clustering/K-means/grid分类
计算机与自动化引用本文复制引用
欧阳浩,陈波,王萌,黄镇谨..基于网格的二次K-means聚类算法[J].广西工学院学报,2012,23(1):24-27,33,5.基金项目
广西科技攻关计划项目 （）
广西工学院博士基金项目（院科博11Z05）资助 （）
