改进的k-prototypes算法及应用

The Improred K-prototypes Algorithmic and Applies

文中提出了一种改进的k-prototypes算法,该算法可以解决具有数值和分类混合类型数据的聚类问题,将它应用于对某高校网站的Web服务器日志进行数据分析,发现有意义的信息,建立规则库,并验证了算法的有效性。

This paper discusses an improved k-prototypes algorithm,which can solve the clustering problem of numerical value and classified, mixed-type data. Applying the algorithm to the Web server log on a university website for data analysis, this paper found some meaningful information to set up the rule hank and verified the effectiveness of the algorithm.