报告题目：Newsvendor under Mean-Variance Ambiguity and Misspecification

时间：2024年11月25日上午10：00-11:00

地点：中关村校区主楼216

报告人：陈植

报告人简介：

Zhi Chen is an Associate Research Fellow at CUHK Shenzhen Research Institute and an Assistant Professor in the CUHK Business School, the Chinese University of Hong Kong. His research interests include (1) developing models and designing algorithms for decision-making under uncertainty with different levels of data availability as well as applications in business, economics, finance, and operations; (2) how to compete or cooperate in joint activities such as resource allocation and risk management. He worked in the City University of Hong Kong and received a Research Excellence Award from the College of Business.

报告内容简介：

Consider a newsvendor problem with an unknown demand distribution. When addressing the issue of distributional uncertainty, we distinguish ambiguity under which the newsvendor does not differentiate demand distributions of common distributional characteristics (e.g., mean and variance) and misspecification under which such characteristics might be misspecified (due to, e.g., estimation error and/or distribution shift). Focusing on the popular mean-variance ambiguity set and optimal-transport cost for the misspecification, we show that the decision criterion of misspecification aversion possesses insightful interpretations as distributional transforms and convex risk measures. We derive the closed-form optimal order quantity that generalizes the solution of the seminal Scarf model under only ambiguity aversion. We quantify the finite-sample performance guarantee, which consists of two parts: the in-sample optimal value and the out-of-sample effect of misspecification that can be decoupled into estimation error and distribution shift. This theoretically justifies the necessity of incorporating misspecification aversion in a non-stationary environment, which is also well demonstrated in our experiments with real-world retailing data. Our framework can be extended to consider multiple products, distributional characteristics specified via optimal transport, and misspecification measured by phi-divergence.

（承办：管理工程系、科研与学术交流中心、中国运筹学会数据科学与运筹智能分会（筹））