CAM Colloquium: Wen Sun (CS, Cornell) - Generalization and robustness in offline reinforcement learning


Frank H. T. Rhodes Hall 655


Offline Reinforcement Learning (RL) is a learning paradigm where the RL agent only learns from a pre-collected static dataset and cannot further interact with the environment anymore. Offline RL is a promising approach for safety-critical applications where randomized exploration and experimentation are not safe. In this talk, we study offline RL in large scale problems with rich function approximation. In the first part of the talk, we will study the generalization property in offline RL and we will give a general model-based offline RL algorithm that provably generalizes in large scale Markov Decision Processes. Our approach is also robust in the sense that as long as there is a high-quality policy whose traces are covered by the offline data, our algorithm will find it. In the second part of the talk, we consider the offline Imitation Learning (IL) setting where the RL agent has an additional set of high-quality expert demonstrations. In this setting, we give an IL algorithm that learns with polynomial sample complexity and achieves start-of-art performance in standard continuous control robotics benchmark.

Wen Sun is an assistant professor in the Computer Science Department at Cornell. Prior to Cornell, he was a postdoc researcher at Microsoft Research, New York City. He completed his PhD at the Robotics Institute at Carnegie Mellon University in 2019. Much of his research is about designing algorithms for efficient decision making under uncertainty, understanding exploration and exploitation in Reinforcement Learning, and how to leverage expert demonstrations to further speed up learning.