Ethan W. Che

PhD Student, Columbia Business School
Decision, Risk, and Operations


I am a PhD student at Columbia Business School in the Decision, Risk, and Operations division. I am advised by Jing Dong and Hongseok Namkoong.

My research is on structured reinforcement learning for large-scale decision-making problems in operations research. In particular, I've developed new algorithmic learning approaches in the domains of (1) bandit exploration and adaptive learning and (2) scheduling in discrete-event systems. My work uses tools from control theory, RL, gradient estimation, and stochastic modeling.

I worked as an ML engineer intern at Instacart in Spring 2023 working on causal machine learning. In Summer 2024, I was an ML Research intern at Netflix working on bandit exploration and uncertainty quantification.

eche25 (at) gsb.columbia.edu      Google Scholar     


Under Review

How Well do LLMs Compress Their Own Chain-of-Thought?
A Token Complexity Approach
Ayeong Lee, Ethan Che, and Tianyi Peng
arXiv:2503.01141 [cs.CL], 2025
arXiv

Differentiable Discrete Event Simulation for Queuing Network Control
Ethan Che, Jing Dong, and Hongseok Namkoong
arXiv:2409.03740 [cs.LG], 2024
Preliminary version in ICML 2023 Differentiable Almost Everything Workshop
arXiv workshop

Mathematical Programming for Adaptive Experiments
Ethan Che, Daniel R. Jiang, Hongseok Namkoong, Jimmy Wang
arXiv:2408.04570 [cs.LG], 2024
Preliminary version in ESIF Economics and AI+ML Meeting 2024 (33% acceptance rate)
Selected for presentation at MIT Conference on Digital Experimentation (CODE), 2024
arXiv workshop

Adaptive Experimentation at Scale: A Computational Framework for Flexible Batches
Ethan Che and Hongseok Namkoong
arXiv:2303.11582 [cs.LG], 2023
Major Revision at Operations Research.
arXiv

Stochastic Gradient Descent with Adaptive Data

Ethan Che, Jing Dong, and Xin T. Tong
arXiv:2410.01195 [cs.LG], 2024
Major Revision at Operations Research.
arXiv

AExGym: Benchmarks and Environments for Adaptive Experimentation
Jimmy Wang, Ethan Che, Daniel R. Jiang, and Hongseok Namkoong
arXiv:2408.04531 [cs.LG], 2024
arXiv


Work in Progress

Exploration Sizing via Model-Predictive Control
Ethan Che, Hakan Ceylan, James McInerney, and Nathan Kallus
Selected for presentation at MIT Conference on Digital Experimentation (CODE), 2024
Work in progress.

Discounting in Markov Chain Estimation
Ethan Che and Jing Dong
Work in progress.
pdf


Publications

QGym: Scalable Simulation and Benchmarking of Queuing Network Controllers
Haozhe Chen*, Ang Li*, Ethan Che*, Tianyi Peng, Jing Dong, Hongseok Namkoong
NeurIPS 2024 Datasets and Benchmarks Track
arXiv

Robustly Optimal Auction Design under Mean Constraints
Ethan Che
arXiv:1911.07103 [econ.TH], 2022
EC '22: Proceedings of the 23rd ACM Conference on Economics and Computation
arXiv conference


Talks


Teaching

  • [TA] MBA Core: Managerial Statistics
  • [TA] EMBA Core: Managerial Statistics
  • [TA] MBA Core: Operations Management
  • [TA] MBA Elective: Technology Breakthroughs
  • [TA] PhD Core: Introduction to Econometrics and Statistical Inference I