Sequential decision-making: theory and algorithms
Theory and algorithms for reinforcement learning and stochastic optimization under structural information such as predictions, factorization, and constraints.
Learning, optimization, control, and energy systems
Chenbei Lu
Eric and Wendy Schmidt AI Postdoctoral Fellow, AI for Science Institute, Cornell University
My research develops theoretical and algorithmic foundations for reliable sequential decision-making under uncertainty, with applications in energy systems and AI infrastructure.
CurrentAI for Science Institute, Cornell University
MethodsReinforcement learning · stochastic optimization
SystemsPower systems · storage · AI data centers
About me
Postdoctoral fellow at the AI for Science Institute.
I am an Eric and Wendy Schmidt AI Postdoctoral Fellow at the AI for Science Institute, Cornell University, working with Prof. Fengqi You. My research develops structure-aware reinforcement learning, stochastic optimization, and safe control methods for large-scale energy and AI infrastructure.
Before joining Cornell, I received my Ph.D. in Computer Science from the Institute for Interdisciplinary Information Sciences at Tsinghua University in 2025, advised by Prof. Chenye Wu and Prof. Ran Duan. I was also a visiting student researcher in Computing and Mathematical Sciences at Caltech, advised by Prof. Adam Wierman. I received my bachelor's degree in Computer Software Engineering from Huazhong University of Science and Technology in 2020.
Research areas
Sequential decision-making under structure and uncertainty.
Learning and optimization for energy systems
Application-driven learning, optimization, and control methods for power-system operation, storage control, dispatch, unit commitment, and demand flexibility.
Energy and AI infrastructure
Models and optimization frameworks for AI data centers, regional power systems, supply headroom, and emerging infrastructure planning questions.
Selected publications
Selected papers.
NeurIPS 2025, Spotlight
Reinforcement Learning with Imperfect Transition Predictions: A Bellman-Jensen Approach
PDFICML 2025
Overcoming the Curse of Dimensionality in Reinforcement Learning Through Approximate Factorization
PDFIEEE Transactions on Smart Grid
Self-Improving Online Storage Control for Stable Wind Power Commitment
PDFIEEE Transactions on Power Systems
Sample-Adaptive Robust Economic Dispatch with Statistical Guarantees
PDFIEEE Control Systems Letters
On the Optimal Deterministic Policy Learning in Chance-Constrained Markov Decision Processes
PDFTeaching
Guest lectures and teaching assistance.
Guest lectures
- Fall 2025: AI for Science, Cornell University
- Fall 2025: AI for Energy Systems, Cornell University
- Spring 2026: Deep Learning, Cornell University
Teaching assistance
- Fall 2021: Combinatorial Mathematics, Tsinghua University
- Fall 2022: AI Research Practice, Yao Class, Tsinghua University
News
Updates.
2025
NeurIPS Spotlight paper on reinforcement learning with imperfect predictions.
2025
Eric and Wendy Schmidt AI Postdoctoral Fellow at the AI for Science Institute, Cornell.
Professional service
Reviewer for NeurIPS, AAAI, IEEE CDC, ACC, PSCC, IEEE SmartGridComm, ACM e-Energy, Applied Energy, IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Transactions on Smart Grid, IEEE Transactions on Power Systems, and related journals.
TPC member for IEEE SmartGridComm, ACM e-Energy poster session.
Contact