Yunbei Xu

Yunbei Xu is a Researcher at Columbia University's Department of Decision, Risk, and Operations (DRO) within the Graduate School of Business. He holds a PhD from Columbia University (2023) and a BS in Pure Mathematics from Peking University (highest honor). His research bridges Artificial Intelligence, Decision Science, and Physics, focusing on sequential learning, bandit algorithms, and statistical theory. He has advised four students and co-authored multiple influential papers in venues like NeurIPS and ICML, earning awards including the ICML Outstanding Paper Award.

• Education: PhD in DRO (Columbia, 2023), Postdoc at MIT's College of Computing (LIDS), BS in Pure Mathematics (Peking University)

Research interests include adversarial learning, robust optimization, and algorithmic design for sequential decision-making. His work emphasizes theoretical guarantees with practical applications in inventory management and stochastic modeling. Recent articles explore lower bound frameworks for bandit algorithms and Bayesian principles in frequentist learning. Teaching roles include instructor for Stochastic Models at NUS and assistant for courses at Columbia.

• Awards: INFORMS Nicholson Prize (1st), Applied Probability Society Finalist (2x)

Active in grants and lab collaborations, Xu's work is implemented in repositories like MAB-code. His academic lineage traces back to mathematical pioneers like Hilbert and Bernoulli, reflecting deep roots in foundational theory.