计算机应用 ›› 2009, Vol. 29 ›› Issue (08): 2250-2252.

• 人工智能 • 上一篇    下一篇

基于决策树和相异度算法的移动通信客户分类方法

陈峰   

  1. 浙江省公安厅
  • 收稿日期:2009-03-12 修回日期:2009-05-12 发布日期:2008-08-01 出版日期:2009-08-01
  • 通讯作者: 陈峰
  • 基金资助:

Classification of mobile communication customers based on decision tree and dissimilarity algorithm

  • Received:2009-03-12 Revised:2009-05-12 Online:2008-08-01 Published:2009-08-01

摘要: 移动通信领域中的客户信息挖掘是数据挖掘和商务智能领域中典型应用之一,具有较高研究意义和商业应用价值。在基于决策树的数据分类算法基础上,采用相异度计算原理进行再次分类,以此对移动通信客户信息中的不同属性进行分析,重点对移动通信客户是否可能成长为高价值客户的分类进行了研究。测试结果表明,该组合分类算法在移动通信客户分类时的平均准确率达到了83.1%。

关键词: 数据挖掘, 商务智能, 决策树, 聚类算法, Data Mining (DM), decision tree, clustering algorithm

Abstract: Customer information mining for mobile communication is one of the typical applications in the field of data mining and business intelligence; therefore, it is meaningful for research and contains high value among business applications. In this paper, the dataset was classified by using decision tree algorithm at first, and then based on the results the dataset was reclassified by using dissimilarity algorithm to analyze the different attributes of the customer information. Especially, the issue of predicting whether a customer would be a highvalued customer in the future was studied. Experimental results show that the average classification accuracy of mobile communication customers is as high as 83.1%.

Key words: business intelligence

中图分类号: