Identification of Blackwell Optimal Policies for Deterministic MDPs - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Identification of Blackwell Optimal Policies for Deterministic MDPs

Résumé

This paper investigates a new learning problem, the identification of Blackwell optimal policies on deterministic MDPs (DMDPs): A learner has to return a Blackwell optimal policy with fixed confidence using a minimal number of queries. First, we characterize the maximal set of DMDPs for which the identification is possible. Then, we focus on the analysis of algorithms based on product-form confidence regions. We minimize the number of queries by efficiently visiting the state-action pairs with respect to the shape of confidence sets. Furthermore, these confidence sets are themselves optimized to achieve better performance. The performance of our method compares to the lower bound up to a factor n 2 in the worst case-where n is the number of states, and constant in certain classes of DMDPs.
Fichier principal
Vignette du fichier
paper.pdf (518.41 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04082136 , version 1 (26-04-2023)

Licence

Paternité

Identifiants

  • HAL Id : hal-04082136 , version 1

Citer

Victor Boone, Bruno Gaujal. Identification of Blackwell Optimal Policies for Deterministic MDPs. AISTATS 2023 - 26th International Conference on Artificial Intelligence and Statistics, Apr 2023, Valencia, Spain. pp.32. ⟨hal-04082136⟩
18 Consultations
17 Téléchargements

Partager

Gmail Facebook X LinkedIn More