| 一种有效k-均值聚类中心的选取方法 |
| |
| 引用本文： | 曹文平. 一种有效k-均值聚类中心的选取方法[J]. 计算机与现代化, 2008, 0(3): 95-97 |
| |
| 作者姓名： | 曹文平 |
| |
| 作者单位： | 襄樊学院电气信息工程系,湖北,襄樊,441003 |
| |
| 摘 要： | 基于k-均值算法的思想和关键技术，本文对于k-均值算法中的初始点的选取进行了深入的研究，提出了一种高性能初始点的选取算法并用实际数据进行测试，通过与常规的随机选取方法的比较，该算法具有更好的性能和健壮性。
|
| 关 键 词： | 聚类 K-均值 初始化 中心 |
| 文章编号： | 1006-2475（2008）03-0095-03 |
| 收稿时间： | 2007-03-20 |
| 修稿时间： | 2007-03-20 |
An Effective Method of Selecting Initial Points for k-means Clustering |
| |
| CAO Wen-ping. An Effective Method of Selecting Initial Points for k-means Clustering[J]. Computer and Modernization, 2008, 0(3): 95-97 |
| |
| Authors: | CAO Wen-ping |
| |
| Affiliation: | CAO Wen-ping ( Dept. of Electrical ＆ Information Engineering of Xiangfan University, Xiangfan 441003, China) |
| |
| Abstract: | This paper provides firstly the idea and core technique of k-means clustering,and then focus on selecting the centriod of k-means clustering.Depending on the reseach about initialization deeply,it presents a high quality approach that used to select the centriod.Using the methods to test the algorithm and compare with the random method,it concludes that our method has the high quality and robustness. |
| |
| Keywords: | clustering k-means initialization centriod |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
