主题:  An Online Mirror Descent Learning Algorithm for Multiproduct Inventory Systems

 多品类库存系统的在线镜像下降学习算法

时间: 2026年4月23日 晚上20：00

地点: 腾讯会议 304-467-830

主讲人: Cong Shi 教授

Bio: Cong Shi is a Professor in the Department of Management at Miami Herbert Business School. His research centers on designing and analyzing efficient algorithms for operations management, with main applications in supply chain management, revenue management, service operations, and human-robot interactions. He received his Ph.D. from MIT ORC in 2012. He has published over 30 papers in top-tier journals, including Operations Research, Management Science, Manufacturing & Service Operations Management, and Production and Operations Management. His work has been recognized by multiple best paper awards from INFORMS and POMS.

Abstract:

We study a canonical inventory control problem: a multiproduct periodic-review lost-sales inventory system with a warehouse capacity constraint. We study this well-researched problem under the lens of demand learning from censored data. Unlike traditional literature, we do not assume that demand distributions are known a priori. Instead, the decision-maker has access only to observed sales data, while the lost-sales quantity remains unobserved. Existing online learning algorithms bear limitations in providing good-quality solutions to inventory systems offering a large variety of products. We employ and innovate mirror descent with cyclic update techniques to address the challenge of high-dimensionality in product menus. We prove theoretically that our algorithm’s regret bound exhibits a logarithmic dependence on the number of products. This constitutes a significant improvement compared to the square-root regret bound established in existing literature. Using empirical data, we implemented our methods to assess their practical merit and expose additional managerial insights. Our numerical study confirms that indeed our methodology produces inventory policies superior to existing state-of-the-art solutions, especially when managing a large menu of products. Drawing from our numerical observations and theory-informed insights, we provide clear guidelines for practical implementation, along with fine-tuning recommendations.