Bayesian Reinforcement Learning: A Survey

Abstract

Bayesian methods for machine learning have been widely investigated, yielding principled methods for incorporating prior information into inference algorithms. In this survey, we provide an in-depth review of the role of Bayesian methods for the reinforcement learning (RL) paradigm. The major incentives for incorporating Bayesian reasoning in RL are: 1) it provides an elegant approach to action-selection (exploration/exploitation) as a function of the uncertainty in learning; and 2) it provides a machinery to incorporate prior knowledge into the algorithms. We first discuss models and methods for Bayesian inference in the simple single-step Bandit model. We then review the extensive recent literature on Bayesian methods for model-based RL, where prior information can be expressed on the parameters of the Markov model. We also present Bayesian methods for model-free RL, where priors are expressed over the value function or policy class. The objective of the paper is to provide a comprehensive survey on Bayesian RL algorithms and their theoretical and empirical properties.

References in corpus (8)

• Further Optimal Regret Bounds for Thompson Sampling
• Model-Based Bayesian Exploration
• A Bayesian Sampling Approach to Exploration in Reinforcement Learning
• An Information-Theoretic Analysis of Thompson Sampling
• Model-Based Bayesian Reinforcement Learning in Large Structured Domains
• Near-Optimal BRL using Optimistic Local Transitions
• Thompson Sampling for Learning Parameterized Markov Decision Processes
• PAC-Bayesian Policy Evaluation for Reinforcement Learning