Offline Reinforcement Learning as Anti-Exploration
Type de document :
Communication dans un congrès avec actes
Titre :
Offline Reinforcement Learning as Anti-Exploration
Auteur(s) :
Rezaeifar, Shideh [Auteur]
Université de Genève = University of Geneva [UNIGE]
Dadashi, Robert [Auteur]
Google Research [Paris]
Vieillard, Nino [Auteur]
Institut Élie Cartan de Lorraine [IECL]
Biology, genetics and statistics [BIGS]
Google Research [Paris]
Hussenot, Léonard [Auteur]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Google Research [Paris]
Bachem, Olivier [Auteur]
Google Research [Zurich]
Pietquin, Olivier [Auteur]
Google Research [Paris]
Geist, Matthieu [Auteur]
Google Research [Paris]
Université de Genève = University of Geneva [UNIGE]
Dadashi, Robert [Auteur]
Google Research [Paris]
Vieillard, Nino [Auteur]
Institut Élie Cartan de Lorraine [IECL]
Biology, genetics and statistics [BIGS]
Google Research [Paris]
Hussenot, Léonard [Auteur]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Scool [Scool]
Google Research [Paris]
Bachem, Olivier [Auteur]
Google Research [Zurich]
Pietquin, Olivier [Auteur]
Google Research [Paris]
Geist, Matthieu [Auteur]
Google Research [Paris]
Titre de la manifestation scientifique :
AAAI 2022 - 36th AAAI Conference on Artificial Intelligence
Organisateur(s) de la manifestation scientifique :
Association for the Advancement of Artificial Intelligence (AAAI)
Ville :
Vancouver
Pays :
Canada
Date de début de la manifestation scientifique :
2022-02-22
Discipline(s) HAL :
Informatique [cs]/Apprentissage [cs.LG]
Résumé en anglais : [en]
Offline Reinforcement Learning (RL) aims at learning an optimal control from a fixed dataset, without interactions with the system. An agent in this setting should avoid selecting actions whose consequences cannot be ...
Lire la suite >Offline Reinforcement Learning (RL) aims at learning an optimal control from a fixed dataset, without interactions with the system. An agent in this setting should avoid selecting actions whose consequences cannot be predicted from the data. This is the converse of exploration in RL, which favors such actions. We thus take inspiration from the literature on bonus-based exploration to design a new offline RL agent. The core idea is to subtract a prediction-based exploration bonus from the reward, instead of adding it for exploration. This allows the policy to stay close to the support of the dataset. We connect this approach to a more common regularization of the learned policy towards the data. Instantiated with a bonus based on the prediction error of a variational autoencoder, we show that our agent is competitive with the state of the art on a set of continuous control locomotion and manipulation tasks.Lire moins >
Lire la suite >Offline Reinforcement Learning (RL) aims at learning an optimal control from a fixed dataset, without interactions with the system. An agent in this setting should avoid selecting actions whose consequences cannot be predicted from the data. This is the converse of exploration in RL, which favors such actions. We thus take inspiration from the literature on bonus-based exploration to design a new offline RL agent. The core idea is to subtract a prediction-based exploration bonus from the reward, instead of adding it for exploration. This allows the policy to stay close to the support of the dataset. We connect this approach to a more common regularization of the learned policy towards the data. Instantiated with a bonus based on the prediction error of a variational autoencoder, we show that our agent is competitive with the state of the art on a set of continuous control locomotion and manipulation tasks.Lire moins >
Langue :
Anglais
Comité de lecture :
Oui
Audience :
Internationale
Vulgarisation :
Non
Collections :
Source :
Fichiers
- http://arxiv.org/pdf/2106.06431
- Accès libre
- Accéder au document
- 2106.06431
- Accès libre
- Accéder au document