摘要
在给出K-means算法流程的基础上,以校园网认证计费系统日志为研究对象,实现一个基于K-means算法的校园网用户行为聚类分析,得到不同特征的用户组,为今后进一步了解用户行为特征、更好制定校园网络策略奠定基础。
The working flow of K- means arithmetic is presented firstly in the paper. And using the network authentication and charge system log as research object, a clustering analysis of network customer＇s action based on K - means arithmetic is implemented. The clustering result discovered some customer＇s groups with varied characters, and established basic for further making different network policy for different customer＇s groups.
出处
《计算技术与自动化》
2007年第1期66-69,共4页
Computing Technology and Automation
基金
国家计委西部大学校园网络工程项目(计投资20012437)
关键词
K—means
聚类分析
用户行为
K- means arithmetic
clustering analysis
customer＇s action
