Regret Bounds for Reinforcement Learning with Policy Advice - Archive ouverte HAL Access content directly
Conference Papers Year : 2013

Regret Bounds for Reinforcement Learning with Policy Advice

(1) , (2, 3) , (1)
1
2
3

Abstract

In some reinforcement learning problems an agent may be provided with a set of input policies, perhaps learned from prior experience or provided by advisors. We present a reinforcement learning with policy advice (RLPA) algorithm which leverages this input set and learns to use the best policy in the set for the reinforcement learning task at hand. We prove that RLPA has a sub-linear regret of $\widetilde O(\sqrt{T})$ relative to the best input policy, and that both this regret and its computational complexity are independent of the size of the state and action space. Our empirical simulations support our theoretical analysis. This suggests RLPA may offer significant advantages in large domains where some prior good policies are provided.
Fichier principal
Vignette du fichier
RLPAcr.pdf (242.17 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-00924021 , version 1 (06-01-2014)

Identifiers

  • HAL Id : hal-00924021 , version 1

Cite

Mohammad Gheshlaghi Azar, Alessandro Lazaric, Emma Brunskill. Regret Bounds for Reinforcement Learning with Policy Advice. ECML/PKDD - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Sep 2013, Prague, Czech Republic. ⟨hal-00924021⟩
297 View
328 Download

Share

Gmail Facebook Twitter LinkedIn More