| 一种基于信息熵的聚类结果评价方法 |
| |
| 引用本文： | 张敞,王园园,赵裕啸,伍章俊.一种基于信息熵的聚类结果评价方法[J].合肥工业大学学报(自然科学版),2011,34(8):1251-1256. |
| |
| 作者姓名： | 张敞 王园园 赵裕啸 伍章俊 |
| |
| 作者单位： | 合肥工业大学管理学院,安徽合肥230009;合肥工业大学过程优化与智能决策教育部重点实验室,安徽合肥230009 |
| |
| 基金项目： | 国家863高技术研究发展计划资助项目(2007AA04Z116); 国家自然科学基金资助项目(70801025); 国家社会科学基金资助项目(10CGL024) |
| |
| 摘 要： | 文章基于信息熵理论,将模糊聚类评价方法和决策树分类算法联系起来,提出从信息熵角度用决策树分类算法来评价聚类算法结果的有效性,从而确定最佳的聚类个数；并将该方法应用到证券行业客户忠诚度分析模型的建立中.实验结果表明,该方法可以明显提高聚类的效果,并且使得聚类结果的可解释性强,具有良好的实用价值.
|
| 关 键 词： | 信息熵 聚类评价 决策树 客户忠诚度 |
A clustering results evaluation method based on information entropy |
| |
| ZHANG Chang,WANG Yuan-yuan,ZHAO Yu-xiao,WU Zhang-jun.A clustering results evaluation method based on information entropy[J].Journal of Hefei University of Technology(Natural Science),2011,34(8):1251-1256. |
| |
| Authors: | ZHANG Chang WANG Yuan-yuan ZHAO Yu-xiao WU Zhang-jun |
| |
| Institution: | ZHANG Chang1,2,WANG Yuan-yuan1,ZHAO Yu-xiao1,WU Zhang-jun1,2(1.School of Management,Hefei University of Technology,Hefei 230009,China,2.Key Laboratory of Process Optimization and Intelligent Decision-making of Ministry of Education,China) |
| |
| Abstract: | In this paper,based on the theory of information entropy,the decision tree classification method is introduced to evaluate the effectiveness of the results of fuzzy clustering algorithm.In this way,the optimal number of clustering can be determined.This evolution method is applied to establishing the customer fidelity model for securities busines.The experiment results show that the method is practical,and can markedly improve the clustering efficiency and make the clustering results more interpretable. |
| |
| Keywords: | information entropy clustering evaluation decision tree customer fidelity |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
