Meta-learning Parameterized Skills

Haotian Fu, Shangqun Yu, Saket Tiwari, Michael Littman, George Konidaris
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:10461-10481, 2023.

Abstract

We propose a novel parameterized skill-learning algorithm that aims to learn transferable parameterized skills and synthesize them into a new action space that supports efficient learning in long-horizon tasks. We propose to leverage off-policy Meta-RL combined with a trajectory-centric smoothness term to learn a set of parameterized skills. Our agent can use these learned skills to construct a three-level hierarchical framework that models a Temporally-extended Parameterized Action Markov Decision Process. We empirically demonstrate that the proposed algorithms enable an agent to solve a set of highly difficult long-horizon (obstacle-course and robot manipulation) tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v202-fu23f, title = {Meta-learning Parameterized Skills}, author = {Fu, Haotian and Yu, Shangqun and Tiwari, Saket and Littman, Michael and Konidaris, George}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {10461--10481}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/fu23f/fu23f.pdf}, url = {https://proceedings.mlr.press/v202/fu23f.html}, abstract = {We propose a novel parameterized skill-learning algorithm that aims to learn transferable parameterized skills and synthesize them into a new action space that supports efficient learning in long-horizon tasks. We propose to leverage off-policy Meta-RL combined with a trajectory-centric smoothness term to learn a set of parameterized skills. Our agent can use these learned skills to construct a three-level hierarchical framework that models a Temporally-extended Parameterized Action Markov Decision Process. We empirically demonstrate that the proposed algorithms enable an agent to solve a set of highly difficult long-horizon (obstacle-course and robot manipulation) tasks.} }
Endnote
%0 Conference Paper %T Meta-learning Parameterized Skills %A Haotian Fu %A Shangqun Yu %A Saket Tiwari %A Michael Littman %A George Konidaris %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-fu23f %I PMLR %P 10461--10481 %U https://proceedings.mlr.press/v202/fu23f.html %V 202 %X We propose a novel parameterized skill-learning algorithm that aims to learn transferable parameterized skills and synthesize them into a new action space that supports efficient learning in long-horizon tasks. We propose to leverage off-policy Meta-RL combined with a trajectory-centric smoothness term to learn a set of parameterized skills. Our agent can use these learned skills to construct a three-level hierarchical framework that models a Temporally-extended Parameterized Action Markov Decision Process. We empirically demonstrate that the proposed algorithms enable an agent to solve a set of highly difficult long-horizon (obstacle-course and robot manipulation) tasks.
APA
Fu, H., Yu, S., Tiwari, S., Littman, M. & Konidaris, G.. (2023). Meta-learning Parameterized Skills. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:10461-10481 Available from https://proceedings.mlr.press/v202/fu23f.html.

Related Material