Skip to Main content Skip to Navigation
Conference papers

Adaptive Batch Size for Safe Policy Gradients

Matteo Papini 1 Matteo Pirotta 2 Marcello Restelli 1
2 SEQUEL - Sequential Learning
Inria Lille - Nord Europe, CRIStAL - Centre de Recherche en Informatique, Signal et Automatique de Lille - UMR 9189
Abstract : Policy gradient methods are among the best Reinforcement Learning (RL) techniques to solve complex control problems. In real-world RL applications, it is common to have a good initial policy whose performance needs to be improved and it may not be acceptable to try bad policies during the learning process. Although several methods for choosing the step size exist, research paid less attention to determine the batch size, that is the number of samples used to estimate the gradient direction for each update of the policy parameters. In this paper, we propose a set of methods to jointly optimize the step and the batch sizes that guarantee (with high probability) to improve the policy performance after each update. Besides providing theoretical guarantees, we show numerical simulations to analyse the behaviour of our methods.
Document type :
Conference papers
Complete list of metadata

Cited literature [24 references]  Display  Hide  Download
Contributor : Alessandro Lazaric Connect in order to contact the contributor
Submitted on : Friday, December 1, 2017 - 12:19:12 PM
Last modification on : Friday, January 7, 2022 - 3:43:48 AM


Files produced by the author(s)


  • HAL Id : hal-01653330, version 1


Matteo Papini, Matteo Pirotta, Marcello Restelli. Adaptive Batch Size for Safe Policy Gradients. The Thirty-first Annual Conference on Neural Information Processing Systems (NIPS), Dec 2017, Long Beach, United States. ⟨hal-01653330⟩



Les métriques sont temporairement indisponibles