News: paper Theoretically Principled Deep RL Acceleration via Nearest Neighbor Function Approximation is accepted by AAAI 2021! Check out the code and slides.

News: Mathematical Reconstruction of Patient-Specific Vascular Networks is accepted for publication on Jan 3, 2021.

I’m a fourth-year undergraduate student at UCLA, majoring in Mathematics of Computation. Currently, I study sample-efficient reinforcement learning (RL) with Dr. Lin Yang. I also work as a research assistant at Center for Vision, Cognition, Learning, and Autonomy (VCLA), advised by Professor Song-Chun Zhu, Professor Ying Nian Wu, and Ph.D. candidate Luyao Yuan.

I’m broadly interested in how machines can acquire greater learning ability with insights from statistics, optimization, and human cognitive skills. In my previous research, I have studied reinforcement learning, machine teaching, and multi-agent collaboration. I would like to explore more about:

• developing new learning frameworks with theoretical support;
• improving the generalizability and efficiency of RL algorithms with optimization/modeling techniques;
• applying machine learning techniques to real-life problems such as robot control.