Preference-based Evolutionary Direct Policy Search - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2013

Preference-based Evolutionary Direct Policy Search

Robert Busa-Fekete
  • Fonction : Auteur
Balázs Szorenyi
  • Fonction : Auteur
Paul Weng
  • Fonction : Auteur
  • PersonId : 952563
Weiwei Cheng
  • Fonction : Auteur
Eyke Hüllermeier
  • Fonction : Auteur

Résumé

We present a novel approach to preference-based reinforcement learning, namely a preference-based variant of a direct policy search method based on evolutionary optimization. The core of our approach is a preference-based racing algorithm that selects the best among a given set of candidate policies with high probability. To this end, the algorithm operates on a suitable ordinal preference structure and only uses pairwise comparisons between sample rollouts of the policies. We present first experimental studies showing that our approach performs well in practice.
Fichier non déposé

Dates et versions

hal-01216088 , version 1 (15-10-2015)

Identifiants

  • HAL Id : hal-01216088 , version 1

Citer

Robert Busa-Fekete, Balázs Szorenyi, Paul Weng, Weiwei Cheng, Eyke Hüllermeier. Preference-based Evolutionary Direct Policy Search. ICRA Autonomous Learning Workshop, May 2013, Karlsruhe, Germany. ⟨hal-01216088⟩
122 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More