A Fitted-Q Algorithm for Budgeted MDPs

Nicolas Carrara (1, 2) , Romain Laroche (3) , Jean-Léon Bouraoui (1) , Tanguy Urvoy (1) , Olivier Pietquin (4)

1 Orange Labs [Lannion]
2 SEQUEL - Sequential Learning
3 Maluuba
4 CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189

We address the problem of budgeted reinforcement learning, in continuous state-space, using a batch of transitions. To this extend, we introduce a novel algorithm called Budgeted Fitted-Q (BFTQ). Benchmarks show that BFTQ performs as well as a regular Fitted-Q algorithm in a continuous 2-D world but also allows one to choose the right amount of budget that fits to a given task without the need of engineering the rewards. We believe that the general principles used to design BFTQ can be applied to extend others classical reinforcement learning algorithms for budgeted oriented applications.

Mots clés: Budgeted-MDP; Fitted-Q; Reinforcement Learning

Domaines: Intelligence artificielle [cs.AI]

Nicolas Carrara, Romain Laroche, Jean-Léon Bouraoui, Tanguy Urvoy, Olivier Pietquin. A Fitted-Q Algorithm for Budgeted MDPs. EWRL 2018 - 14th European workshop on Reinforcement Learning, Oct 2018, Lille, France. ⟨hal-01928092⟩