About me

Hi! I’m Zechen, a researcher in Computer Science at Brown University, working under the guidance of Amy Greenwald and Ronald Parr. Our latest research paper, A Unifying View of Linear Function Approximation in Off-Policy RL Through Matrix Splitting and Preconditioning, is now available. Click here to read it.

My primary research interest are:

• understanding behavior, phenomena, and properties of fundamental Reinforcement Learning(RL) algorithms (e.g., Temporal Difference Learning(TD), Q-learning, Policy Gradient) and their derivatives or inspired variants (e.g., Deep Q-Network (DQN), implicit Q-learning) that demonstrate exceptional practical performance, aiming for mathematically provable insights beyond empirical observations and intuition.

• designing efficient RL algorithms capable of solving industrial or real-world problems

• understanding the behavior and phenomena of fundamental RL algorithms when combined with neural networks (and other nonlinear function approximators)

• connecting reinforcement learning and continual learning