Regret Bounds for Learning State Representations in Reinforcement Learning - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Regret Bounds for Learning State Representations in Reinforcement Learning

Résumé

We consider the problem of online reinforcement learning when several state representations (mapping histories to a discrete state space) are available to the learning agent. At least one of these representations is assumed to induce a Markov decision process (MDP), and the performance of the agent is measured in terms of cumulative regret against the optimal policy giving the highest average reward in this MDP representation. We propose an algorithm (UCB-MS) with O(√ T) regret in any communicating MDP. The regret bound shows that UCB-MS automatically adapts to the Markov model and improves over the currently known best bound of order O(T 2/3).
Fichier principal
Vignette du fichier
9435-regret-bounds-for-learning-state-representations-in-reinforcement-learning.pdf (319.15 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02375715 , version 1 (22-11-2019)

Identifiants

  • HAL Id : hal-02375715 , version 1

Citer

Ronald Ortner, Matteo Pirotta, Ronan Fruit, Alessandro Lazaric, Odalric-Ambrym Maillard. Regret Bounds for Learning State Representations in Reinforcement Learning. Conference on Neural Information Processing Systems, Dec 2019, Vancouver, Canada. ⟨hal-02375715⟩
64 Consultations
167 Téléchargements

Partager

Gmail Facebook X LinkedIn More