Optimistic Posterior Sampling for Reinforcement Learning with Few Samples and Tight Guarantees

Code

• github.com/d-tiapkin/optimistic-psrl-experiments
  OfficialIn papernone★ 1

Abstract

We consider reinforcement learning in an environment modeled by an episodic, finite, stage-dependent Markov decision process of horizon H with S states, and A actions. The performance of an agent is measured by the regret after interacting with the environment for T episodes. We propose an optimistic posterior sampling algorithm for reinforcement learning (OPSRL), a simple variant of posterior sampling that only needs a number of posterior samples logarithmic in H, S, A, and T per state-action pair. For OPSRL we guarantee a high-probability regret bound of order at most O(H^3SAT) ignoring poly(HSAT) terms. The key novel technical ingredient is a new sharp anti-concentration inequality for linear forms which may be of independent interest. Specifically, we extend the normal approximation-based lower bound for Beta distributions by Alfers and Dinges [1984] to Dirichlet distributions. Our bound matches the lower bound of order (H^3SAT), thereby answering the open problems raised by Agrawal and Jia [2017b] for the episodic setting.

Tasks

Reproductions