Budgeted Reinforcement Learning in Continuous State Space

Carrara, Nicolas; Leurent, Edouard; Laroche, Romain; Urvoy, Tanguy; Maillard, Odalric Ambrym; Pietquin, Olivier

Document type :

Communication dans un congrès avec actes

Title :

Budgeted Reinforcement Learning in Continuous State Space

Author(s) :

Carrara, Nicolas [Auteur]
Sequential Learning [SEQUEL]
Leurent, Edouard [Auteur]
RENAULT
Sequential Learning [SEQUEL]
Laroche, Romain [Auteur]
Urvoy, Tanguy [Auteur]
Orange Labs [Lannion]
Maillard, Odalric Ambrym [Auteur]

Sequential Learning [SEQUEL]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]
Pietquin, Olivier [Auteur]
Sequential Learning [SEQUEL]
Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189 [CRIStAL]

Conference title :

Conference on Neural Information Processing Systems

City :

Vancouver

Country :

Canada

Start date of the conference :

2019-12

Journal title :

Advances in Neural Information Processing Systems

HAL domain(s) :

Statistiques [stat]/Machine Learning [stat.ML]

English abstract : [en]

A Budgeted Markov Decision Process (BMDP) is an extension of a Markov Decision Process to critical applications requiring safety constraints. It relies on a notion of risk implemented in the shape of a cost signal constrained ...
Show more >A Budgeted Markov Decision Process (BMDP) is an extension of a Markov Decision Process to critical applications requiring safety constraints. It relies on a notion of risk implemented in the shape of a cost signal constrained to lie below an-adjustable-threshold. So far, BMDPs could only be solved in the case of finite state spaces with known dynamics. This work extends the state-of-the-art to continuous spaces environments and unknown dynamics. We show that the solution to a BMDP is a fixed point of a novel Budgeted Bellman Optimality operator. This observation allows us to introduce natural extensions of Deep Reinforcement Learning algorithms to address large-scale BMDPs. We validate our approach on two simulated applications: spoken dialogue and autonomous driving.Show less >

Language :

Anglais

Peer reviewed article :

Oui

Audience :

Internationale

Popular science :

Non

ANR Project :

BANDITS MANCHOTS POUR SIGNAUX NON-STATIONNAIRES ET STRUCTURES

Collections :

Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL) - UMR 9189

Source :

Harvested from HAL

Files

https://hal.archives-ouvertes.fr/hal-02375727/document
Open access
Access the document

http://arxiv.org/pdf/1903.01004
Open access
Access the document

https://hal.archives-ouvertes.fr/hal-02375727/document
Open access
Access the document

https://hal.archives-ouvertes.fr/hal-02375727/document
Open access
Access the document

document
Open access
Access the document

9128-budgeted-reinforcement-learning-in-continuous-state-space.pdf
Open access
Access the document

1903.01004
Open access
Access the document

document
Open access
Access the document

9128-budgeted-reinforcement-learning-in-continuous-state-space.pdf
Open access
Access the document

Budgeted Reinforcement Learning in Continuous ... BibTeX CSV Excel RIS

Files

Budgeted Reinforcement Learning in Continuous ...

BibTeX

CSV

Excel

RIS