Beyond Value-Function Gaps: Improved Instance-Dependent Regret Bounds for Episodic Reinforcement Learning

07/02/2021
by   Christoph Dann, et al.
6

We provide improved gap-dependent regret bounds for reinforcement learning in finite episodic Markov decision processes. Compared to prior work, our bounds depend on alternative definitions of gaps. These definitions are based on the insight that, in order to achieve a favorable regret, an algorithm does not need to learn how to behave optimally in states that are not reached by an optimal policy. We prove tighter upper regret bounds for optimistic algorithms and accompany them with new information-theoretic lower bounds for a large class of MDPs. Our results show that optimistic algorithms can not achieve the information-theoretic lower bounds even in deterministic MDPs unless there is a unique optimal policy.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/07/2022

Cascaded Gaps: Towards Gap-Dependent Regret for Risk-Sensitive Reinforcement Learning

In this paper, we study gap-dependent regret guarantees for risk-sensiti...
research
05/09/2019

Non-Asymptotic Gap-Dependent Regret Bounds for Tabular MDPs

This paper establishes that optimistic algorithms attain gap-dependent a...
research
05/20/2019

Issues concerning realizability of Blackwell optimal policies in reinforcement learning

N-discount optimality was introduced as a hierarchical form of policy- a...
research
11/19/2020

Online Model Selection for Reinforcement Learning with Function Approximation

Deep reinforcement learning has achieved impressive successes yet often ...
research
01/09/2020

Self-guided Approximate Linear Programs

Approximate linear programs (ALPs) are well-known models based on value ...
research
08/06/2018

Regret Bounds for Reinforcement Learning via Markov Chain Concentration

We give a simple optimistic algorithm for which it is easy to derive reg...
research
07/18/2022

An Information-Theoretic Analysis of Bayesian Reinforcement Learning

Building on the framework introduced by Xu and Raginksy [1] for supervis...

Please sign up or login with your details

Forgot password? Click here to reset