时 间：4月10日上午九点

主 题：Personalized Product Recommendation with Exponential Bandit

地 点：九里校区零号楼0411室

内 容：We adopt a single-agent multi-armed exponential bandit framework to solve a platform's personalized product recommendation problem, where the platform does not know whether the consumer has a demand for a product. The platform updates its belief continually and can switch freely among all products in the assortment before the consumer makes a purchase. We obtain three main results. First, we show that the optimal policy for selecting products follows an index theorem similar to the Gittins index. Under the optimal policy, the platform selects one product each time, and a product is selected at most once. Second, we show that the platform may not regard information acquisition as a last-ditch effort before quitting, which contradicts the existing literature. Third, with a discount factor, we demonstrate a tradeoff between high-revenue products and fast-purchase products. Due to this tradeoff, a product may be selected multiple times, but the number of times is bounded above. Our results can be generalized to the case of multiple dependent/independent sets.

主讲人简介：刘昊昱博士现任澳门城市大学商学院助理教授。他分别于2017年和2020年获得香港科技大学商学院运营管理硕士学位和博士学位。他采用混合方法（分析模型以及实证和实验工具）来解决营销和运营界面的问题，特别关注新兴技术和消费者行为问题。他文章发表在《Omega》和《Service Science》等国际高水平期刊上，此外还有多篇文章在Manufacturing and Service Operations Management 和Production and Operations Management 第二轮审稿中。

主办：437必赢会员中心 运营管理创新团队

承办：服务科学与创新四川省重点实验室

欢迎公司师生学习交流！