Timezone: »

 
Poster
Efficient Bayes-Adaptive Reinforcement Learning using Sample-Based Search
Arthur Guez · David Silver · Peter Dayan

Thu Dec 06 02:00 PM -- 12:00 AM (PST) @ Harrah’s Special Events Center 2nd Floor #None

Bayesian model-based reinforcement learning is a formally elegant approach to learning optimal behaviour under model uncertainty, trading off exploration and exploitation in an ideal way. Unfortunately, finding the resulting Bayes-optimal policies is notoriously taxing, since the search space becomes enormous. In this paper we introduce a tractable, sample-based method for approximate Bayes-optimal planning which exploits Monte-Carlo tree search. Our approach outperformed prior Bayesian model-based RL algorithms by a significant margin on several well-known benchmark problems -- because it avoids expensive applications of Bayes rule within the search tree by lazily sampling models from the current beliefs. We illustrate the advantages of our approach by showing it working in an infinite state space domain which is qualitatively out of reach of almost all previous work in Bayesian exploration.

Author Information

Arthur Guez (DeepMind)
David Silver (DeepMind)
Peter Dayan (Gatsby Unit, UCL)

I am Director of the Gatsby Computational Neuroscience Unit at University College London. I studied mathematics at the University of Cambridge and then did a PhD at the University of Edinburgh, specialising in associative memory and reinforcement learning. I did postdocs with Terry Sejnowski at the Salk Institute and Geoff Hinton at the University of Toronto, then became an Assistant Professor in Brain and Cognitive Science at the Massachusetts Institute of Technology before moving to UCL.

More from the Same Authors