电信客户流失量估计组合模型的仿真研究
|
摘要
研究电信客户流失问题,电信客户流失数据具有模糊性和非线性,单一算法仅能对模糊性或非线性进行预测,为提高电信客户流失估计准确率,提出了一种电信客户流失组合估计模型。首先对客户属性进行清洗并进行离散化处理,接着使用粗糙集方法对离散属性进行约简,刻画电信客户流失数据的模糊性;然后遗传算法优化支持向量机对电信客流失非线性进行描述,建立电信客户流失估计模型。仿真结果表明,粗糙集与支持向量机相融合模型克服单一粗糙集算法或支持向量机存在的缺陷,提高电信客户流失估计模型的估计准确率,可为电信客户管理优化设计提供依据。
Aiming at the problems of the complexity of telecommunication customer loss and to improve prediction accuracy rate,this paper proposed a new telecom customer churn prediction model based on rough set theory and support vector machine(SVM).Firstly,customer properties were cleaned and discretely processed;then discrete attributes were reduced by using rough set method,and the parameters of SVM were optimized by genetic algorithm to obtain the optimal parameters.Finally,the performance of model was tested.The experimental results show that the hit rate,covering rate,accuracy rate and lift coefficient of the proposed model are higher than those of the reference model.
Aiming at the problems of the complexity of telecommunication customer loss and to improve prediction accuracy rate,this paper proposed a new telecom customer churn prediction model based on rough set theory and support vector machine(SVM).Firstly,customer properties were cleaned and discretely processed;then discrete attributes were reduced by using rough set method,and the parameters of SVM were optimized by genetic algorithm to obtain the optimal parameters.Finally,the performance of model was tested.The experimental results show that the hit rate,covering rate,accuracy rate and lift coefficient of the proposed model are higher than those of the reference model.
引文
[1]吴基传.世界电信业分析与思考[M].北京:新华出版社,2002-10.
[2]黄海.数据挖掘在电信客户流失估计中的应用[D].北京:北京邮电大学,2006-3.
[3]钱苏丽,何建敏,王纯麟.基于改进支持向量机的电信客户流失估计模型[J].管理科学,2007,20(1):54-58.
[4]M C Mozer.Predicting Subscriber Dissatisfaction and ImprovingRetention in the W ireless Telecommunications Industry[J].IEEETrans on Neural Networks,2000,11(3):690-696.
[5]Lian Yan.Predicting Customer Behavior in Telecommunications[J].IEEE Intelligent Systems,2004,19(2):50-58.
[6]罗彬,等.基于粗糙集理论-神经网络-蜂群算法集成的客户流失研究[J].管理学报,2011,8(2):265-272.
[7]陈坚,等.基于粗糙集的公路客运量估计[J].重庆交通大学学报(自然科学版),2009,28(6):1071-1074.
[8]夏国恩,金炜东.基于支持向量机的客户流失估计模型[J].系统工程理论与实践,2008,(1):71-77.
[9]杨庆德,李建珍.利用遗传算法辅助设计人工神经网络的权值和域值[J].天津大学学报,2003,36(6):782-786.
[10]Y Wei,D N Jutla,S C Sivakumar.A churn-strategy alignmentmodel for managers in mobile telecom communication[C].Net-works and Services Research Conference,2005:48-53.
[11]夏国恩,邵培基.改进的支持向量分类机在客户流失估计中的应用[J].计算机应用研究,2009,26(6):2044-2046.
[12]刘勇健,刘义建,张伯友.基于粗糙集—支持向量机的震害估计模型及应用[J].地震研究,2008,31(3):289-295.
[2]黄海.数据挖掘在电信客户流失估计中的应用[D].北京:北京邮电大学,2006-3.
[3]钱苏丽,何建敏,王纯麟.基于改进支持向量机的电信客户流失估计模型[J].管理科学,2007,20(1):54-58.
[4]M C Mozer.Predicting Subscriber Dissatisfaction and ImprovingRetention in the W ireless Telecommunications Industry[J].IEEETrans on Neural Networks,2000,11(3):690-696.
[5]Lian Yan.Predicting Customer Behavior in Telecommunications[J].IEEE Intelligent Systems,2004,19(2):50-58.
[6]罗彬,等.基于粗糙集理论-神经网络-蜂群算法集成的客户流失研究[J].管理学报,2011,8(2):265-272.
[7]陈坚,等.基于粗糙集的公路客运量估计[J].重庆交通大学学报(自然科学版),2009,28(6):1071-1074.
[8]夏国恩,金炜东.基于支持向量机的客户流失估计模型[J].系统工程理论与实践,2008,(1):71-77.
[9]杨庆德,李建珍.利用遗传算法辅助设计人工神经网络的权值和域值[J].天津大学学报,2003,36(6):782-786.
[10]Y Wei,D N Jutla,S C Sivakumar.A churn-strategy alignmentmodel for managers in mobile telecom communication[C].Net-works and Services Research Conference,2005:48-53.
[11]夏国恩,邵培基.改进的支持向量分类机在客户流失估计中的应用[J].计算机应用研究,2009,26(6):2044-2046.
[12]刘勇健,刘义建,张伯友.基于粗糙集—支持向量机的震害估计模型及应用[J].地震研究,2008,31(3):289-295.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心