详细信息

电子商务中基于Q学习的动态交叉销售方法    

Q-learning-based dynamic cross-selling approach in e-commerce

文献类型:期刊文献

中文题名:电子商务中基于Q学习的动态交叉销售方法

英文题名:Q-learning-based dynamic cross-selling approach in e-commerce

作者:程岩[1]

机构:[1]华东理工大学商学院,上海200237

年份:2008

卷号:11

期号:3

起止页码:106

中文期刊名:管理科学学报

外文期刊名:Journal of Management Sciences in China

收录:人大复印报刊资料;CSTPCD;;北大核心:【北大核心2004】;CSSCI:【CSSCI2008_2009】;

语种:中文

中文关键词:电子商务;交叉销售;Q-学习;关联规则

外文关键词:electronic commerce ; cross-selling ; Q-learning; association rule

摘要:动态交叉销售是电子商务中的一种新型营销手段.在已知关联规则和商品库存水平的情况下,要研究两个决策问题:(1)如何选择交叉销售的商品组合(2)如何为商品组合确定合理的价格,从而使经销商获得最大收益.首先将动态交叉销售映射为事件驱动的马氏决策过程模型,其次结合关联规则理论提出了知识驱动的 Q-学习算法,K-Q-learning,用该算法来求解动态交叉销售问题具有较高的效率和效用.
Dynamic cross-selling is a novel marketing technique in electronic commerce setting. Given association rules and the level of inventory, two issues are analyzed: (1) how to select packaging complements and (2) how to price product packages to maximize profits. First the cross-selling problem was formulated as a event-driven markov decision process model, then a knowledge-driven Q-learning algorithm based on association rule, K-Q-learning, was proposed, which has high efficiency and effectiveness in solving the dynamic crossselling problem.

参考文献:

正在载入数据...

版权所有©华东理工大学 重庆维普资讯有限公司 渝B2-20050021-7 
渝公网安备 50019002500408号 违法和不良信息举报中心