The third invited lecture for the new verification and validation of models course is on Monday (Oct 19, 2020) at 7:20 PM EST. This week, Dr. Bilal Kartal, Founder in Residence at Entrepreneur First, will give a lecture titled “Safer Deep Reinforcement Learning and Auxiliary Tasks”. If you’re interested in joining the lecture, please email me (see the flyer below) to receive the Zoom invitation link.
Safe reinforcement learning has many variants, and it is still an open research problem. In this talk, I describe different auxiliary tasks that improve learning and focus on using action guidance through a non-expert demonstrator to avoid catastrophic events in a domain with sparse, delayed, and deceptive rewards: the previously proposed multi-agent benchmark of Pommerman. I present a framework where a non-expert simulated demonstrator, e.g., planning algorithms such as Monte Carlo tree search with a small number of rollouts, can be integrated into asynchronous distributed deep reinforcement learning methods. Compared to vanilla deep RL algorithms, our proposed methods both learn faster and converge to better policies on a two-player mini version of the Pommerman game and Atari games.