A Fitted-Q Algorithm for Budgeted MDPs - Archive ouverte HAL Accéder directement au contenu
Autre Publication Scientifique Année : 2018

A Fitted-Q Algorithm for Budgeted MDPs

Nicolas Carrara
  • Fonction : Auteur
  • PersonId : 1036158
Romain Laroche
  • Fonction : Auteur
  • PersonId : 1012067
Tanguy Urvoy
  • Fonction : Auteur
  • PersonId : 933382

Résumé

We address the problem of bud-geted/constrained reinforcement learning in continuous state-space using a batch of transitions. For this purpose, we introduce a novel algorithm called Budgeted Fitted-Q (BFTQ). We carry out some preliminary benchmarks on a continuous 2-D world. They show that BFTQ performs as well as a penalized Fitted-Q algorithm while also allowing ones to adapt the trained policy on-the-fly for a given amount of budget and without the need of engineering the reward penalties. We believe that the general principles used to design BFTQ could be used to extend others classical reinforcement learning algorithms to budget-oriented applications.
Fichier principal
Vignette du fichier
ncarrara-saferl-uai-2018.pdf (7.38 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01867353 , version 1 (04-09-2018)

Identifiants

  • HAL Id : hal-01867353 , version 1

Citer

Nicolas Carrara, Romain Laroche, Jean-Léon Bouraoui, Tanguy Urvoy, Olivier Pietquin. A Fitted-Q Algorithm for Budgeted MDPs. 2018. ⟨hal-01867353⟩
326 Consultations
232 Téléchargements

Partager

Gmail Facebook X LinkedIn More