Developing Autonomous Agents to Learn and Plan in Simulation and the Real-World

Par

Glen Berseth

University of California, Berkeley.

Résumé:

While humans plan and solve tasks with ease, simulated and robotics agents struggle to reproduce the same fidelity, robustness and skill. For example, humans can grow to perform incredible gymnastics, prove that black holes exist, and produce works of art, all starting from the same base learning system. If we can design an agent with a similar learning capability, the agent can acquire skills through experience, without the need for expertly constructed planning systems or supervision. In this talk, I present a series of developments on current reinforcement learning methods in the area of long-term planning and learning autonomously without human guidance or supervision. I show how modularity and policy reuse can be used to address challenges in long-horizon planning. Still, learning using current RL methods requires types of supervision that are easy to come by in simulation but are expensive in open and real worlds. I will discuss how to develop more versatile learning agents that do not require expensive or unrealistic data constraints. Last, in an effort to create agents that learn general-purpose skills I present an objective for learning sophisticated control over the environment.

Biographie :

Glen Berseth is a Postdoctoral Researcher with Berkeley Artificial Intelligence Research (BAIR) working in the Robotic AI & Learning (RAIL) lab with Sergey Levine. Dr. Berseth completed his NSERC-awarded Ph.D. in Computer Science at the University of British Columbia in 2019, where he worked with Michiel van de Panne. He received his BSc degree in Computer Science from York University in 2012 and an MSc from York University under the supervision of Petros Faloutsos in 2014. He has published in a range of top venues in machine learning, robotics and computer animation. The goal of his research is to develop systems that can learn and act in the world intelligently. Much of this work has focused on developing deep learning and reinforcement learning methods to solve complex, high-dimensional perception and planning problems.