Bayesian Optimization with Automatic Prior Selection for Data-Efficient Direct Policy Search - Archive ouverte HAL Access content directly
Conference Papers Year :

Bayesian Optimization with Automatic Prior Selection for Data-Efficient Direct Policy Search

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

Abstract

One of the most interesting features of Bayesian optimization for direct policy search is that it can leverage priors (e.g., from simulation or from previous tasks) to accelerate learning on a robot. In this paper, we are interested in situations for which several priors exist but we do not know in advance which one fits best the current situation. We tackle this problem by introducing a novel acquisition function, called Most Likely Expected Improvement (MLEI), that combines the likelihood of the priors and the expected improvement. We evaluate this new acquisition function on a transfer learning task for a 5-DOF planar arm and on a possibly damaged, 6-legged robot that has to learn to walk on flat ground and on stairs, with priors corresponding to different stairs and different kinds of damages. Our results show that MLEI effectively identifies and exploits the priors, even when there is no obvious match between the current situations and the priors.
Fichier principal
Vignette du fichier
1709.06919.pdf (3.4 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01768279 , version 1 (17-04-2018)

Identifiers

Cite

Rémi Pautrat, Konstantinos Chatzilygeroudis, Jean-Baptiste Mouret. Bayesian Optimization with Automatic Prior Selection for Data-Efficient Direct Policy Search. ICRA 2018 - International Conference on Robotics and Automation, May 2018, Brisbane, Australia. ⟨hal-01768279⟩
175 View
283 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More