Hello, I’m Kai

Hello! I am a Ph.D. candidate in Applied Mathematics and Statistics on the Statistics Track at Stony Brook University. My work centers on reinforcement learning, statistical modeling, simulation, and sequential decision-making.

This website is both my academic home and a small personal archive: a place for my research and teaching, the courses and books that shaped me, and a few ideas that I simply find delightful.

View my research Teaching Download my CV Download my résumé GitHub LinkedIn

Selected work

Cumulative tic-tac-toe

I am the first author of a peer-reviewed Stats article combining combinatorial game theory with temporal-difference reinforcement learning.

Read the article · View the official implementation

RL-QESA

I am a co-first author of an AI for Math Workshop at ICML 2025 paper on reinforcement-learning-guided temperature control for simulated annealing.

Read the OpenReview paper · View the publication page

Current work

My current work includes a co-first-author manuscript comparing six reinforcement learning algorithms across expanded blackjack action spaces and a co-authored manuscript on regional sea-level prediction. Further details are available in my CV.

Teaching

I have taught or supported courses in mathematical statistics, statistical learning, time series, and statistical computing. I also enjoy mentoring students as they learn how to turn questions into careful analyses and clear scientific communication.

See my teaching and mentoring record

A few ideas I enjoy

“Block what you can, randomize what you cannot.” — Box, Hunter, and Hunter

I also have a soft spot for Simpson’s paradox, the birthday problem, Occam’s razor, and the reminder that bad inputs rarely produce good analysis.

Visit the less formal side of the site

Why “Kai’s Square”?

The name is a small statistics joke: Kai and chi sound alike in English, and χ² is never far away in a statistics program. The longer story is on my About page.