PhD student / MIT ORC & LIDS

Jack
Zhang.

Making better decisions
in an uncertain world.

I study reinforcement learning and partially observable systems, with a focus on what we can learn from limited information—and how to make reliable decisions with it.

Explore my research View publications
FIG. 01LEARNING FROM WHAT WE OBSERVE
A sketch of partially observable decision-making Hidden states influence observations. Observations inform a decision. Curved paths connect the parts of the system. HIDDEN STATE OBSERVATION ACTION unknownsignalschoice THE CENTRAL QUESTION What can the data tell us?
✳ REINFORCEMENT LEARNING / POMDPs / INFERENCE
CURRENTLYPhD Student at MIT ORC & LIDS
RESEARCH AREASRL · POMDPs · Statistical Learning
LOCATIONCambridge, Massachusetts

01 / RESEARCH

Understanding the hidden.
Learning the possible.

My work sits at the intersection of reinforcement learning, optimization, and statistical inference. I'm particularly interested in settings where an agent never sees the true underlying state of the world.

01

Learning under partial observability

How can we recover hidden dynamics and plan effectively when observations reveal only part of the system?

POMDPsSystem identification
02

Robust decision-making

How do we design policies that remain dependable when models are imperfect and environments are uncertain?

Robust RLOptimization
03

Statistical foundations of learning

What conditions make learning possible, and what guarantees can we establish from finite trajectories?

IdentifiabilitySample complexity

Ongoing Research

Work in Progress

Active Probing for POMDP Identification

Jack Zhang, Saurabh Amin, Patrick Jaillet

Work in Progress

Strategic Learning in Partially Observable Online Marketplaces

Jack Zhang, Owen Shen, Saurabh Amin, Patrick Jaillet

02 / PUBLICATIONS AND PREPRINTS

Selected work.

A home for papers and preprints on learning, inference, and sequential decision-making.

2026
NeurIPS 2026 · Accepted

From Likelihood Convergence to Parameter Convergence in POMDPs

Jack Zhang, Saurabh Amin, Jiawei Zhang, Patrick Jaillet

2026
Mathematics of Operations Research · Submitted

Approximate Solution of Infinite-Horizon Risk-Sensitive Markov Decision Processes

Jack Zhang, Mehdi Hemmati, Saumya Sinha, Andrew J. Schaefer

2026
INFORMS Journal on Computing · Accepted

Partitioning a Graph into Low-Diameter Clusters

Jack Zhang , Lucas Silveira, Hamidreza Validi, Logan Smith, Austin Buchanan, Illya V. Hicks

[Paper]

2024
Optimization Letters · Accepted

Linear-Size Formulations for Connected Planar Graph Partitioning and Political Districting

Jack Zhang , Hamidreza Validi, Austin Buchanan, Illya V. Hicks

[Journal] [PDF] [Code]

More publication details will be added as they become available.

03 / ABOUT

A little
about me.

I'm a PhD student in Operations Research at the Massachusetts Institute of Technology.

I study the mathematical foundations of learning and decision-making, especially in partially observable settings. I enjoy questions that connect theory with the realities of sequential data: what a model can identify, what an agent can know, and how to act when uncertainty cannot be eliminated.

AFFILIATION
Massachusetts Institute of TechnologyPhD · Operations Research
RESEARCH INTERESTS
Reinforcement learning and inferencePartially observable systems, robust optimization, statistical learning

04 / GET IN TOUCH

Let's connect.

I'm always interested in thoughtful conversations about research, collaborations, and new ideas.

Contact details and profile links will be added here.

Cambridge, MassachusettsKeep exploring ↑ Back to top