Xuefeng Gao

Associate Professor
Department of Systems Engineering and Engineering Management
The Chinese University of Hong Kong

Room 606, William M. W. Mong Engineering Building
The Chinese University of Hong Kong
Shatin, N.T., Hong Kong
Email: xfgao AT se.cuhk.edu.hk
Phone: (852)3943-8242

Educations

• Ph.D. in Operations Research, School of Industrial and Systems Engineering, Georgia Institute of Technology, 2013.
• B.S. in Mathematics, School of Mathematical Sciences, Peking University, 2008. 

Research Interests

• Applied Probability 
• Online Learning and Reinforcement Learning
• Generative Diffusion Models
• Queueing Theory
• Algorithmic Trading

Publications and Preprints

• Remanufacturing inventory system with demand-dependent returns: optimality analysis and approximations (with Chenxi Sun, Zhijie Tao and Sean Zhou)
  Working paper. (Link)


• Reward-directed score-based diffusion models via q-Learning (with Jiale Zha and Xunyu Zhou)
  Working paper. (Link)


• Reinforcement learning for intensity control: an application to choice-based network revenue management (with Huiling Meng and Ningyuan Chen)
  Working paper. (Link)


• Reinforcement learning for jump-diffusions, with financial applications (with Lingfei Li and Xunyu Zhou)
  Working paper. (Link) 


• Reinforcement learning for continuous-time optimal execution: Actor-Critic algorithm and error analysis (with Boyu Wang and Lingfei Li)
  Working paper. (Link) 


• Regret bounds for episodic risk-sensitive linear quadratic regulator (with Wenhao Xu and Xuedong He)
  Working paper. (Link)  


• No Algorithmic Collusion in two-player blindfolded game with Thompson Sampling (with Ningyuan Chen and Yi Xiong)
  Working paper. (Link) 


• Convergence analysis for general probability flow ODEs of Diffusion models in Wasserstein distances (with Lingjiong Zhu)
  Working paper. (Link) 


• Wasserstein convergence guarantees for a general class of score-based generative models (with Hoang M. Nguyen and Lingjiong Zhu)
  Working paper. (Link) 


• Non-convex stochastic optimization via non-reversible stochastic gradient Langevin dynamics (with Yuanhan Hu, Xiaoyu Wang, Mert Gurbuzbalaban and Lingjiong Zhu)
  Working paper. (Link)


• Square-root regret bounds for continuous-time episodic Markov decision processes (with Xunyu Zhou)
  Mathematics of Operations Research, accepted, 2024. (Link) 


• Logarithmic regret bounds for continuous-time average-reward Markov decision processes (with Xunyu Zhou)
  SIAM Journal on Control and Optimization, accepted, 2024. (Link) 


• Regret bounds for Markov decision processes with recursive optimized certainty equivalents (with Wenhao Xu and Xuedong He)
  ICML, accepted, 2023. (Link) 


• Asymptotically optimal control of make-to-stock systems (with Junfei Huang)
  Mathematics of Operations Research, accepted, 2023. (Link)


• Asymptotically optimal control of omnichannel service systems with pick-up guarantees (with Junfei Huang and Jiheng Zhang)
  Operations Research, accepted, 2023. (published version)


• A data-driven deep learning approach for options market making (with Qianhui Lai and Lingfei Li)
  Quantitative Finance, 23(5), 777-797, 2023. (published version)


• Sublinear regret for learning POMDPs (with Ningyuan Chen, Yi Xiong and Xiang Zhou)
  Production and Operations Management, Vol. 31, No. 9, p.3491 - 3504, 2022. (published version)


• Debiasing samples from online learning using bootstrap (with Ningyuan Chen and Yi Xiong)
  Conference on Artificial Intelligence and Statistics (AISTATS) 2022, oral presentation (top 3% of submissions). (Published version)


• State-dependent temperature control for Langevin diffusions (with Zuoquan Xu and Xunyu Zhou)
  SIAM Journal on Control and Optimization, Vol. 60, 3, 2022. (published version) 


• Global Convergence of Stochastic Gradient Hamiltonian Monte Carlo for non-convex stochastic optimization: Non-asymptotic performance bounds and momentum-based acceleration (with Mert Gurbuzbalaban and Lingjiong Zhu)
  Operations Research, Vol. 70, No. 5, p 2931 - 2947, 2022. (Download; published version)


• Order Scoring, Bandit Learning and Order Cancellations (with Tianrun Xu)
  Journal of Economic Dynamics and Control, Vol. 134, 104287, 2022. (published version)  


• Regime Switching Bandits (with Ningyuan Chen, Yi Xiong and Xiang Zhou)
  NeurIPS, Vol. 34, p. 4542 - 4554, 2021. (Published version)


• Decentralized stochastic gradient Langevin dynamics and Hamiltonian Monte Carlo (with Yuanhan Hu, Mert Gurbuzbalaban and Lingjiong Zhu)
  Journal of Machine Learning Research, Vol. 22, No.1, p. 10804 - 10872, 2021. (Download)


• Breaking reversibility accelerates Langevin Dynamics for global non-convex optimization (with Mert Gurbuzbalaban and Lingjiong Zhu)
  NeurIPS, Vol 33, p. 17850 - 17862, 2020. (Download)


• Optimal market making in the presence of latency (with Yunhan Wang).
  Quantitative Finance, Vol. 20, No. 9, p. 1495-1521, 2020. (Download; published version)


• Optimal order execution using hidden orders (with Yuanyuan Chen and Duan Li).
  Journal of Economic Dynamics and Control, 94, p.89-116, 2018. (published version)


• Transform analysis for Hawkes processes with applications in dark pool trading (with Xiang Zhou and Lingjiong Zhu).
  Quantitative Finance, Vol. 18, No. 2, p. 265-282, 2018. (Download; published version)


• A functional central limit theorem for stationary Hawkes processes and its applications to infinite-server queues (with Lingjiong Zhu).
  Queueing Systems, 90(1-2), p.161-206, 2018 (Download; published version)


• Limit theorems for Markovian Hawkes processes with a large initial intensity (with Lingjiong Zhu).
  Stochastic Processes and Their Applications, 128 (11), p. 3807-3839, 2018. (Download; published version)


• Large deviations and applications for Markovian Hawkes processes with a large initial intensity (with Lingjiong Zhu).
  Bernoulli, 24(4A), p. 2875-2905, 2018 (Download; published version)


• Hydrodynamic limit of order book dynamics (with Shijie Deng)
  Probability in the Engineering and Informational Sciences, 32(1), 96-125, 2018. (Download; published version)  - Internet supplement (PDF)


• Validity of heavy-traffic steady-state approximations in many-server queues with abandonment (with Jim Dai and Ton Dieker).
  Queueing Systems, 78, p. 1-29, 2014. (Download; published version)


• Sensitivity analysis for diffusion processes constrained to an orthant (with Ton Dieker).
  The Annals of Applied Probability, 24, p. 1918-1945, 2014. (Download; published version)


• Positive recurrence of piecewise Ornstein-Uhlenbeck processes and common quadratic Lyapunov functions (with Ton Dieker).
  The Annals of Applied Probability, 23, p. 1291-1317, 2013. (Download; published version)


• Stochastic optimal control for a general class of dynamic resource allocation problems (with Yingdong Lu, Mayank Sharma, Mark Squillante and Joost Bosman) .
  ACM SIGMETRICS Performance Evaluation Review, 41(2), p. 3-14, 2014. (Download)

Grants and Awards

• PI, GRF grant: "Regret bounds for risk sensitive linear quadratic control", 2024-2027.
• PI, GRF grant: "Online learning in games and algorithmic collusion", 2023-2026, with Ningyuan Chen (Co-I).
• PI, GRF grant: "Logarithmic regret bounds for learning in continuous time Markov decision processes", 2022-2025.
• PI, GRF grant: "Temperature control for Langevin diffusions", 2022-2024, with Xunyu Zhou (Co-I).
• PI, GRF grant: "Multi-armed bandits with regime switching rewards", 2021-2023, with Ningyuan Chen (Co-I).
• PI, GRF grant: "Optimal market making for large-tick liquid stocks", 2018-2020.
• PI, ECS grant: "Spread crossing and order placement in limit order markets", 2016-2019.
• PI, CUHK Direct Allocation Grant: "Asymptotic analysis in limit order markets", 2015-2017.
• PI, CUHK Direct Allocation Grant: "Order fill probability in algorithmic trading", 2014-2016.
• Finalist in the 2011 INFORMS Junior Faculty Interest Group (JFIG) paper competition

Teaching

• SEEM5580: Advanced Stochastic Models (PhD)
• ENGG2430/2450: Probability and Statistics for Engineers (Undergraduate)
• SEEM3580: Risk Analysis for Financial Engineering (Undergraduate)
• SEEM3570: Stochastic Models (Undergraduate)
• SEEM5870: Computational Finance (MSc)

Professional Activities

• Reviewer for journals: Annals of Applied Probability, Finance and Stochastics, Mathematics of Operations Research, Operations Research, Stochastic Systems, Stochastic Processes and their Applications etc.

Prospective students

• I am looking for self-motivated PhD students who are interested in Stochastic modelling, Reinforcement learning, Generative AI/diffusion models, and their applications in operations research and financial engineering.