I work with students at the intersection of digital finance, mathematical modeling, and AI. My goal is to help students build both the theoretical foundations and practical skills needed for research or industry in quantitative finance and data science — through problem-driven projects with real-world relevance.

I am looking for students and collaborators with interests in stochastic control / mathematical finance, machine learning / deep learning, quantitative research or algorithmic trading, or decentralized finance (DeFi). Most projects are in active early development, so there is genuine room to shape the direction. Feel free to reach out if any of the areas below interest you.

Current Research Groups

Automated Market Makers (AMMs): Focused on the mathematical and computational design of decentralized exchange protocols — including stochastic control models, game-theoretic liquidity provision, and JAX-based simulation. Projects range from analytical theory to reinforcement learning for fee optimization.

Prediction Markets & High-Frequency Dynamics: Focused on building learning systems for short-term binary option markets and sports betting markets on platforms like Polymarket — using world models (JEPA), temporal graph neural networks, and self-supervised representations trained directly on raw market data.

Graduated Theses

Wei-Ru Chen (M.S., co-advised with Yuki Chino, June 2024)

A Study of AMM Mechanisms and Liquidity Provider Rewards. (Thesis)

Undergraduate Research Projects

2025–2026 NCTS Research Program

Hai-Ching Shih — Design a better AMM.

2024 NCTS Summer Research Program

Co-advisor: A. Christian Silva. Students: Tzu-Ling Hsieh, Guan Ting Lu, Tai-Jun Lai, Mu-Chian Lin. Optimization Methods in Quantitative Trading. (Notes)

2024 Project: Predictive Modeling for Vehicle Allocation

Students: Shih-Bin Chen, Jie-Hong Lai, Jia-Wei Liao, Yuan-Hong Lin. Predictive Modeling for Optimized Vehicle Allocation and Routing in Postal Delivery. Best Application Award & GLORY Data Application Innovation Award (Postal Big Data Competition, 2024). (Notes)

2023–2024 NCTS Research Program

Co-advisor: Chang-Ye Tu. Students: De-Jin Huang, Chen-Chung Yang. Deep Hedging — applying deep learning to derivative hedging strategies. (Notes)

2023 Independent Project: Time Series Generation

Hua-Hsuan Shih — Sig-Wasserstein GANs for Time Series Generation. Now at Columbia MSFE, working in the energy market industry. (Repository)