Neural Fitted Actor-Critic - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2016

Neural Fitted Actor-Critic

Matthieu Zimmer
Yann Boniface
  • Fonction : Auteur
  • PersonId : 778629
  • IdRef : 140410805
Alain Dutech

Résumé

A novel reinforcement learning algorithm that deals with both continuous state and action spaces is proposed. Domain knowledge requirements are kept minimal by using non-linear estimators and since the algorithm does not need prior trajectories or known goal states. The new actor-critic algorithm is on-policy, offline and model-free. It considers discrete time, stationary policies, and maximizes the discounted sum of rewards. Experimental results on two common environments, showing the good performance of the proposed algorithm, are presented.
Fichier principal
Vignette du fichier
main.pdf (410.48 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01350651 , version 1 (23-08-2016)

Identifiants

  • HAL Id : hal-01350651 , version 1

Citer

Matthieu Zimmer, Yann Boniface, Alain Dutech. Neural Fitted Actor-Critic. ESANN 2016 - Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, Apr 2016, Bruges, Belgium. ⟨hal-01350651⟩
331 Consultations
330 Téléchargements

Partager

Gmail Facebook X LinkedIn More