| 移动通信用户信用度的预测评估 |
| |
| 引用本文： | 徐继生,李湛,王道恒,刘鹏.移动通信用户信用度的预测评估[J].电信科学,2002,18(8):17-20. |
| |
| 作者姓名： | 徐继生 李湛 王道恒 刘鹏 |
| |
| 作者单位： | 1. 武汉大学电子信息学院,武汉,430072
2. 广东移动通信公司珠海分公司,珠海,519015 |
| |
| 基金项目： | 广东移动通信公司资助项目，， |
| |
| 摘 要： | 本文提出了一种以遗传算法和多维判别模型为核心的移动用户信用度预测评估方法并给出了试应用的结果。对以广东移动珠海分公司全球通用户为例进行的信用度预测评估，准确率接近90%，误判率和漏判率均低于15%。这种方法对于各级移动通信公司都是极有意义的。如推广应用，必将带来很大的社会和经济效益。
|
| 关 键 词： | 移动通信 多维判别模型 信用度 预测评估 遗传算法 误判率 漏判率 |
| 修稿时间： | 2002年7月4日 |
An Approach to the Forecast and Evaluation of the Credits for Mobile Clients |
| |
| Xu Jisheng,Li Zhan,Wang Daoheng,Liu Peng.An Approach to the Forecast and Evaluation of the Credits for Mobile Clients[J].Telecommunications Science,2002,18(8):17-20. |
| |
| Authors: | Xu Jisheng Li Zhan Wang Daoheng Liu Peng |
| |
| Abstract: | An approach to the evaluation of the credits for mobile clients is proposed in this paper and a result of a testing application in Zhuhai Branch, GMCC. The essential method for the credit_evaluation used in this study is an organic coalescence of client behavior analysis and genetic algorithm. The precision of the evaluation and forecasting of the client's credit is near 90%. At the same time, the judge_by_mistake rate and the leaving_out rate are both under 15%. This study is very significant to the mobile communication corporation at all levels. The popularization of the techniques and the result will produce great benefits of both society and economy. |
| |
| Keywords: | credit_forecast and evaluation genetic algorithm judge_by_mistake rate leaving_out rate |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
