[1]何云 李辉 姚能坚 赵榕生.改进K-means算法实现移动通信行为特征分析[J].计算机技术与发展,2011,(06):63-65.
　HE Yun,LI Hui,YAO Neng-jian,et al.Application of Improved K-Means Algorithm in Mobile Communication Behavioral Characteristic Analysis[J].,2011,(06):63-65.
点击复制

改进K-means算法实现移动通信行为特征分析()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2011年06期
页码:
63-65
栏目:
智能、算法、系统工程
出版日期:
1900-01-01

文章信息/Info

Title:
Application of Improved K-Means Algorithm in Mobile Communication Behavioral Characteristic Analysis
文章编号:
1673-629X（2011）06-0063-03
作者:
何云 李辉 姚能坚 赵榕生
广州军区空军指挥自动化工作站
Author(s):
HE Yun; LI Hui; YAO Neng-jian; ZHAO Rong-sheng
Command Automation Office,Guangzhou Military Region Air Force
关键词:
客户细分; K-means; 影响因子
Keywords:
customer segmentation;  K-means;  influencing factor
分类号:
TP301.6
文献标志码:
A
摘要:
K-means算法被广泛用于客户细分聚类应用研究,客户细分对移动通信行业具有重要的商业价值。但变量的量纲、维度、聚类数、初始聚点等参数的计算是影响K-means算法聚类应用效果的重要因子。在基于K-means算法移动通信行为特征分析系统的实现过程中,分别从特征维度选择、变量量纲统一、聚类数K值与初始聚点的确定等四个方面改进算法的上述影响参数的计算方法,并利用经验加权的方式使算法与主观经验结合。研究结果表明改进K-means算法对移动通信特征分析客户聚类有效
Abstract:
K-means algorithm is widely used to customer segmentation clustering application research,customer segmentation of mobile communications has important commercial value.But dimensionunit,dimension of variable,cluster numbers,initial centroids,etc.calculation of these parameters is important factor of influencing K-means algorithm cluster application result.Based on K-means algorithm mobile communication behavior characteristic analysis process of implementing,respectively from the characteristic dimensions selection,variable dimensionunit unity,cluster number K value and initial centroids determination four aspects,improve the determination of the above algorithm affects parameters calculation method,utilize experience weighting way to make algorithm bind with subjective experience.The result of study indicates that according to behavioral characteristic analysis improving K-means algorithm will subdivide the cluster to the mobile communication customer effectively

相似文献/References:

[1]范新 沈闻 丁泉勋 沈洁.基于正例和未标文档的半监督分类研究[J].计算机技术与发展,2009,(06):58.
　FAN Xin,SHEN Wen,DING Quan-xun,et al.Research on Semi- Supervised Classification Based on Positive and Unlabeled Text Document[J].,2009,(06):58.
[2]王园园 倪志伟 赵裕啸 伍章俊.基于决策树的模糊聚类评价算法及其应用[J].计算机技术与发展,2009,(09):232.
　WANG Yuan-yuan,NI Zhi-wei,ZHAO Yu-xiao,et al.Fuzzy Clustering Evaluation Algorithm Based on Decision Tree and Application[J].,2009,(06):232.
[3]李若鹏 李翔 林祥 李建华.基于DK算法的互联网热点主动发现研究与实现[J].计算机技术与发展,2008,(09):1.
　LI Ruo-peng,LI Xiang,LIN Xiang,et al.Discovering Information Hotspots on Initiative over Internet Based on DK Clustering Algorithm[J].,2008,(06):1.
[4]朱云贺 张春海 张博.基于数据分段的K-means的优化研究[J].计算机技术与发展,2010,(11):130.
　ZHU Yun-he,ZHANG Chun-hai,ZHANG Bo.Optimizing Research on K-means Based on Data Partition[J].,2010,(06):130.
[5]黎银环,张剑.改进的 K-means 算法在入侵检测中的应用[J].计算机技术与发展,2013,(01):165.
　LI Yin-huan,ZHANG Jian.Application of Improved K-means Clustering Algorithm in Intrusion Detection[J].,2013,(06):165.
[6]李四海,满自斌.自适应特征权重的K-means聚类算法[J].计算机技术与发展,2013,(06):98.
　LI Si-hai[],MAN Zi-bin[].K-means Clustering Algorithm Based on Adaptive Feature Weighted[J].,2013,(06):98.
[7]耿永政,陈坚.结合图论的JSEG彩色图像分割算法[J].计算机技术与发展,2014,24(05):15.
　GENG Yong-zheng,CHEN Jian.JSEG Color Image Segmentation Algorithm Combining Graph Theory[J].,2014,24(06):15.
[8]陈斌,苏一丹,黄山. 基于KM-SMOTE和随机森林的不平衡数据分类[J].计算机技术与发展,2015,25(09):17.
　CHEN Bin,SU Yi-dan,HUANG Shan. Classification of Imbalance Data Based on KM-SMOTE Algorithm and Random Forest[J].,2015,25(06):17.
[9]叶雨晴,邱晓晖.基于SIFT 与 K-means 的图像复制粘贴篡改检测[J].计算机技术与发展,2018,28(06):121.[doi:10.3969/ j. issn.1673-629X.2018.06.027]
　YE Yu-qing,QIU Xiao-hui.Copy-move Forgery Detection Based on SIFT and K-means Clustering[J].,2018,28(06):121.[doi:10.3969/ j. issn.1673-629X.2018.06.027]
[10]胡磊,蔡红霞,俞涛. 双重聚类的协同过滤算法在智能家居中的应用[J].计算机技术与发展,2017,27(02):100.
　HU Lei,CAI Hong-xia,YU Tao. Application of Collaborative Filtering Recommendation Based on Double Clustering in Smart Home System[J].,2017,27(06):100.

备注/Memo

备注/Memo:
广空预研项目（GK2009BE0102）何云（1974-），男，湖北天门人，硕士，工程师，CCF会员，研究方向为通信技术、网络安全
更新日期/Last Update: 1900-01-01