南京理工大学学报（自然科学版）2019，Vol.43Issue(1)：108-114,7.DOI:10.14177/j.cnki.32-1397n.2019.43.01.015
基于在线序列优化极限学习机的电子商务客户流失量预测模型
Predictions model of customer churn in E-commerce based on online sequential optimization extreme learning machine
摘要
Abstract
In order to improve the customer churn prediction accuracy of E-commerce customer,and single machine model cannot effectively predict customer churn of massive E-commerce customers, this paper proposes a novel prediction model of customer churn in E-commerce based on online sequential optimization extreme learning machine. Firstly, the Map/Reduce model of cloud computing is used to segment the amount of customer churn in E-commerce,and multiple training subsets are obtained; secondly extreme learning machine is used to model each training subset of E-commerce customer churn,and the prediction results of training subsets are combined to get the final forecast results of customer churn in E-commerce;at last the validity of E-commerce customer churn prediction model is tested by example. The results show that the proposed model improves the prediction accuracy of customer churn in E-commerce,and the training time of E-commerce customer churn modeling has greatly reduced,improving the churn prediction speed of E-commerce customers.关键词
电子商务/客户流失量/云计算处理技术/预测模型/极限学习机Key words
E-commerce/ customer churn/ cloud computing technology/ prediction model/ extreme Learning Machine分类
信息技术与安全科学引用本文复制引用
杨力..基于在线序列优化极限学习机的电子商务客户流失量预测模型[J].南京理工大学学报（自然科学版）,2019,43(1):108-114,7.基金项目
国家自然科学基金(71871082) （71871082）
安徽省高校人文社会科学研究重大项目(SK2016SD15) （SK2016SD15）
安徽省高校优秀青年人才支持计划重点项目(gxyqZD2016456) （gxyqZD2016456）
