Adaptive Batch Size for Safe Policy Gradients - Archive ouverte HAL Access content directly
Conference Papers Year :

Adaptive Batch Size for Safe Policy Gradients

(1) , (2) , (1)
1
2

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.
Fichier principal
Vignette du fichier
6950-adaptive-batch-size-for-safe-policy-gradients.pdf (434.19 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01653330 , version 1 (01-12-2017)

Identifiers

  • HAL Id : hal-01653330 , version 1

Cite

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⟩
209 View
99 Download

Share

Gmail Facebook Twitter LinkedIn More