I-Ta Hsieh

I am a Ph.D. researcher at Brown University standing at the convergence of Computational Materials Science, Applied Mathematics, and Scientific AI. My research focuses on developing rigorous mathematical frameworks and data-driven algorithms to accelerate materials discovery.

I specialize in integrating first-principles theory and molecular dynamics with advanced ML architectures, utilizing Physics-Informed Neural Networks (PINNs) to solve complex differential equations governing defect kinetics. I also develop Machine Learning Force Fields (MLFFs) to bridge the scale gap between quantum accuracy and molecular dynamics, and employ Reinforcement Learning (RL) agents for efficient sampling of high-dimensional energy landscapes.

By synergizing mathematical modeling with high-throughput computational chemistry, I aim to elucidate fundamental mechanisms in solid-state and hybrid systems, driving the next wave of AI-accelerated scientific innovation.

What I Do

Computational Materials Science

• Machine learning and data science for identifying chemical and mechanical properties in materials

• Developing new methods for computing mechanical, thermal, chemical, and transport properties of materials.

• Multi-scale modeling: Ab initio calculation, molecular dynamics simulation, Monte Carlo simulation, and finite difference.

Applied Math and Computer Science

• Physics-Informed Neural Networks (PINNs).

• The non-linear dynamics and the geometry in the neural network.

• High performance computation: parallel computing, quantum computing algorithm.

• Information security for web server.