12月8日讲座预告丨冯一凡：消费者偏好的鲁棒学习方法
数据科学与管理工程学系学术讲座
【主讲题目】Robust Learning of Consumer Preferences
【主讲人】冯一凡，Assistant Professor，NUS Business School
讲座信息
时间：2021年12月8日（周三）上午09:30 –10:30
参与方式：线上腾讯会议（ID : 574 920 388，密码：202112）
主讲人简介
Yifan Feng is an Assistant Professor at the Department of Analytics and Operations (DAO), NUS Business School. Broadly, he is interested in prescriptive analytics for platforms and markets. Specifically, he is interested in stochastic modeling with applications to information aggregation and acquisition, experimentation, and demand fulfillment. Prior to NUS, Yifan obtained his Ph.D. degree from the University of Chicago Booth School of Business.
讲座摘要
A company wants to identify the most preferred product out of a finite set of alternatives when consumer preferences are unknown. It is able to sample consumer preferences by sequentially showing different subsets of products to different consumers and asking them to report their top preferences within the displayed set. The company aims to design a display policy that minimizes the expected number of samples needed to identify the top-ranked product with high probability.
Using a sequential hypothesis testing framework, we prove an instance-specific lower bound on the sample complexity of any policy that identifies the top-ranked version within a given (probabilistic) confidence. We also propose a robust formulation of the company's problem and derive a sampling policy, which is both worst-case asymptotically optimal and intuitive to implement. Roughly speaking, it is a delicate balance among display sets to maximize the informativeness of choices made by the consumers.（This is joint work with René Caldentey and Christopher T. Ryan.）
信息来源：浙大管院数据科学与管理工程学系
今日编辑/排版：段婷
欢迎关注浙江大学管理学院官方微信