Xiaohui Tu

Ph.D. Candidate in Management Science

Department of Decision Sciences
HEC Montréal & Mila - Quebec AI Institute

Portrait photo

I am a Ph.D. candidate in Management Science at HEC Montréal, fortunate to work with Prof. Erick Delage and Prof. Yossiri Adulyasak. I am a student member of GERAD and Mila, and expect to complete my Ph.D. in May 2027. Previously, I received an M.Sc. in Business Analytics from HEC Montréal, working with Prof. Yossiri Adulyasak and Prof. Jean-François Cordeau, and a bachelor's degree in Management Science & Engineering from Nankai University. During my undergraduate studies, I was an exchange student at Korea University.

My research lies at the intersection of operations research and reinforcement learning. I study how shared resources can be allocated efficiently and fairly over time, and how the structure of a decision problem can make large-scale policy optimization tractable.

Doctoral research

My doctoral research studies how to allocate shared resources fairly and efficiently in sequential decision-making problems. Using weakly coupled Markov decision processes as a modeling framework, I investigate how fairness objectives interact with system structure and shape allocation policies. My work develops structural reductions and computational methods that exploit symmetry, aggregate individual states and actions, and retain decision-relevant information about past outcomes. The broader goal is to make fairness-aware decision-making computationally tractable while preserving the information needed to guide resource allocation.

Research interests

Optimization · Weakly coupled Markov decision processes · Reinforcement learning
Fair resource allocation · Transportation