| 电信客户流失的组合预测模型 |
| |
| 引用本文： | 余路,.电信客户流失的组合预测模型[J].华侨大学学报(自然科学版),2016,0(5):637-640. |
| |
| 作者姓名： | 余路 |
| |
| 作者单位： | 1. 西南大学 计算机与信息科技学院, 重庆 北碚 400715; 2. 重庆涪陵广播电视大学 教务处, 重庆 涪陵 408000 |
| |
| 摘 要： | 针对电信行业客户流失的问题,设计基于决策树C5.0、BP神经网络及 Logistic 回归算法的组合预测模型,并对某电信企业进行客户流失预测.预测结果表明:与单一客户流失预测模型相比,组合预测模型命中准确率高,预测效果好,更能直观地显示出流失客户的基本特征.
|
| 关 键 词： | 客户流失 预测模型 电信企业 决策树C5.0 BP神经网络 Logistic回归算法 |
Combination Forecasting Model of Customer Churns in Telecom Industry |
| |
| YU Lu,' target="_blank" rel="external">.Combination Forecasting Model of Customer Churns in Telecom Industry[J].Journal of Huaqiao University(Natural Science),2016,0(5):637-640. |
| |
| Authors: | YU Lu " target="_blank">' target="_blank" rel="external"> |
| |
| Institution: | 1. School of Computer and Information Science, Southwestern University, Chongqing 400715, China; 2. Teaching Affair Office, Chongqing Fuling Radio and television University, Chongqing 408000, China |
| |
| Abstract: | According to telecommunication customer churn problem, the forecasting model based on decision tree C5.0, BP(back-propagation)neural network and logistic regression algorithm combination is designed, and according to orecasting of the customer churns in some telecom companies, the accuracy is higher and prediction effect is good in combination forecasting model compared to a single customer churn prediction model. It shows the basic features of the customer churn more directly. |
| |
| Keywords: | customer churn forecasting model telecom industry decision tree C5 0 back-propagation neural network logistic regression algorithm |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《华侨大学学报(自然科学版)》浏览原始摘要信息 |
| 点击此处可从《华侨大学学报(自然科学版)》下载免费的PDF全文 |
