基于一种改进的K-means算法的校园网用户行为分析研究  

Research of customer behavior analysis in campus network based on an improved K-means clustering algorithm

作　　者：皇甫大鹏[1] 陈平[1] 王兴建[1] 

机构地区：[1]北京师范大学信息网络中心,北京100875

出　　处：《广西大学学报（自然科学版）》2011年第A01期69-72,共4页Journal of Guangxi University（Natural Science Edition）

摘　　要：随着师生对网络流量需求的迅速增长,如何合理的增加用户的流量,既能满足用户的基本上网需求,又能合理利用出口带宽。本文基于一种改进的K-means聚类算法,以校园网用户的行为特征为对象进行聚类,为优化用户计费策略和合理的分配出口带宽提供了数据依据,为校园网用户个性化需求方面提供理论依据。With the rapid growth for network demand of the teachers and students, how to increase the user＇s flow, not only can meet the demand of user＇s basic network flow, but also can reasonablely utilize the export bandwidth. In this paper, based on an improved K-means clustering algorithm, the behavior characteristics of the campus network user were clustered and analyzed, in order to provide a data basis for optimizing the user＇s billing policy, a reasonable allocation of the export bandwidth and the personalized demand of campus network users.

关 键 词：用户行为 聚类算法 K-MEANS 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]