January 12
@
12:00 PM
–
1:00 PM
Assortment and inventory decisions lie at the core of supply chain and retail operations. In practice, these decisions face two fundamental challenges arising from complex customer choice behavior. First, customers often purchase complementary products across categories, which makes category-level decisions interdependent. Second, inventory is limited and customers arrive over time, so product availability changes dynamically as items stock out. Most work in choice-based operations has focused on single-category and static settings, while research addressing these two challenges remains relatively limited. Existing approaches often either oversimplify customer preferences in choice modeling or rely on algorithms that are not tractable in large-scale settings. In this talk, I will present two projects that address these challenges. The first proposes a Markovian framework to model cross-category complementarity, supporting scalable estimation and joint assortment optimization. The second introduces a unified algorithmic framework for dynamic assortment and inventory optimization under MNL choice, with provable guarantees in both personalized and non-personalized settings. Together, these works offer scalable tools for decision-making in complex, data-driven supply chain environments.
Bio: Shuo Sun is a PhD candidate in Industrial Engineering and Operations Research at the University of California, Berkeley. Her research focuses on modeling and algorithm design for supply chain and revenue management using optimization and machine learning. Her work has received several recognitions, including the INFORMS Daniel H. Wagner Prize and a finalist distinction in the INFORMS RMP Jeff McGill Best Student Paper Award. She has publications in leading conferences and papers published or under revision at leading operations and analytics journals, as well as industry experience at Amazon and JD.com on retail operations problems.