Optimism in Reinforcement Learning and Kullback-Leibler Divergence - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2010

Optimism in Reinforcement Learning and Kullback-Leibler Divergence

Résumé

We consider model-based reinforcement learning in finite Markov De- cision Processes (MDPs), focussing on so-called optimistic strategies. In MDPs, optimism can be implemented by carrying out extended value it- erations under a constraint of consistency with the estimated model tran- sition probabilities. The UCRL2 algorithm by Auer, Jaksch and Ortner (2009), which follows this strategy, has recently been shown to guarantee near-optimal regret bounds. In this paper, we strongly argue in favor of using the Kullback-Leibler (KL) divergence for this purpose. By studying the linear maximization problem under KL constraints, we provide an ef- ficient algorithm, termed KL-UCRL, for solving KL-optimistic extended value iteration. Using recent deviation bounds on the KL divergence, we prove that KL-UCRL provides the same guarantees as UCRL2 in terms of regret. However, numerical experiments on classical benchmarks show a significantly improved behavior, particularly when the MDP has reduced connectivity. To support this observation, we provide elements of com- parison between the two algorithms based on geometric considerations.
Fichier principal
Vignette du fichier
KLModelBasedHal.pdf (303.75 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00476116 , version 1 (23-04-2010)
hal-00476116 , version 2 (17-06-2010)
hal-00476116 , version 3 (12-10-2010)

Identifiants

Citer

Sarah Filippi, Olivier Cappé, Aurélien Garivier. Optimism in Reinforcement Learning and Kullback-Leibler Divergence. Communication, Control, and Computing (Allerton), 2010 48th Annual Allerton Conference on, Sep 2010, Monticello (Illinois), United States. pp.115 - 122, ⟨10.1109/ALLERTON.2010.5706896⟩. ⟨hal-00476116v3⟩
220 Consultations
1052 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More