| 用LDA Boosting算法进行客户流失预测 |
| |
| 引用本文： | 应维云,蔺楠,谢雅雅,李秀.用LDA Boosting算法进行客户流失预测[J].数理统计与管理,2010,29(3). |
| |
| 作者姓名： | 应维云 蔺楠 谢雅雅 李秀 |
| |
| 作者单位： | 1. 上海财经大学,上海,200433
2. 清华大学国家CIMS工程研究中心,北京,100084 |
| |
| 基金项目： | 国家自然科学基金(70671059) |
| |
| 摘 要： | 本文提出一种LDA boost(Linear Discriminant Analysis boost)分类方法,该算法能有效利用样本的所有特征,并且能够从高维特征空间里提取并组合优化出最具有判别能力的低维特征,使得样本类间离散度和类内离散度的比值最大,从而不会产生过度学习,大大提高算法效率。该算法有效性在某商业银行的客户流失预测过程的真实数据集中得到了验证。与其他同类算法,如人工神经网络、决策树、支持向量机等运算结果相比,该方法可以显著提高运算精度。同时,LDAboosting与其他boosting算法相比,也具有显著的优越性。
|
| 关 键 词： | 客户流失 数据挖掘 预测 |
Research on the LDA Boosting in Customer Churn Prediction |
| |
| YING Wei-yun,LIN Nan,XIE Ya-ya,LI Xiu.Research on the LDA Boosting in Customer Churn Prediction[J].Application of Statistics and Management,2010,29(3). |
| |
| Authors: | YING Wei-yun LIN Nan XIE Ya-ya LI Xiu |
| |
| Affiliation: | YING Wei-yun~1 LIN Nan~1 XIE Ya-ya~2 LI Xiu~2 (1.Shanghai University of Finance & Economics,Shanghai 200433,China,2.National CIMS Engineering & Research Center,Tsinghua University,Beijing 100084,China) |
| |
| Abstract: | In this paper,a novel classification algorithm called LDA boosting is proposed to predict the customer churn.This algorithm can effectively extract and assemble the most discriminative low-dimensional features from all the features in the high-dimensional feature space of the samples,which make the maximal ratio of intra-class dispersion and inter-class dispersion,thus will not produce exhaustive search,and will greatly improve the learning efficiency.The effectiveness of the proposed algorithm is validated... |
| |
| Keywords: | customer churn data mining prediction |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
