（毕业设计论文）《决策树模型在客户分类中的应用》》.doc

PAGE 、 毕业设计（论文） 题 目 决策树模型在客户分类中 的应用 系 （院） 经济与管理系 专 业 信息管理与信息系统 班 级 2008级1班 学生姓名 学 号 2008060142 指导教师 职 称 讲师 二〇一三年六月 滨州学院本科毕业设计（论文） PAGE I 决策树模型在客户分类中的应用 摘 要 数据挖掘技术，尤其是决策树模型分类方法，在对企业客户资源进行客户分类时具有重要作用和意义。本文介绍了数据挖掘及决策树的具体表述定义，详细阐述了决策树的核心算法ID3算法，同时对于ID3算法在实际应用中所带来的误差和取值偏向问题进行了优化，并提出了优化算法。该算法在选取分支属性时，选取了信息增益度最大的属性，取代了ID3算法所依据的信息增益的标准。另外，在对数值连续型的数据进行离散化时，本文用分类准确率代替了原来的信息增益，简化了计算过程。并且利用ID3算法及改进的算法应用到山东领先生物工程公司的客户资源，建立了决策树分类模型，对客户进行了分类。为该领域的数据挖掘提供了合理的挖掘模式，并得出有效的分析结论。 关键词：数据挖掘；决策树；客户分类；信息增益度 The Application of the Decision Tree Model in Customer Classification Abstract The data mining technology, especially the classification method of decision tree model, plays an important and significant role in customer classification of enterprise’s customer resources. This paper introduces the data mining and decision tree’s specific definition of expression and expounds the decision tree algorithm of the ID3 algorithm. This paper also puts forward the optimization algorithm for the problem of the value approach of ID3 algorithm and the errors caused by ID3 algorithm in the actual application to improve ID3 algorithm. The optimization algorithm is proposed to select the information gain’s biggest attribute as a branch of attribute to replace the ID3 algorithm of the information gain standards. In addition, this article replaces the original information gain with the classification accuracy when splitting the binary data making the calculation greatly simplified. And we apply ID3 algorithm and optimization algorithm to a customer resources management of Ling Xian Biotechnology Company in Shan Dong province which is based on decision tree classification model and provide the reasonable mining mode and the effective analysis conclusion in the field of data mining. Key words: Data Mining; Decision Tree; Customer Classification; Information Gain Degree 目 录 TOC \o 1-3 \h \z \u HYPERLINK \l _Toc326786763 引言 1 HYPERLINK \l _Toc326786764 第一章 数据挖掘及决策树方法