The Max-Min Formulation of Multi-Objective Reinforcement Learning: From Theory to a Model-Free Algorithm

Giseung Park, Woohyeon Byeon, Seongmin Kim, Elad Havakuk, Amir Leshem, Youngchul Sung
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:39616-39642, 2024.

Abstract

In this paper, we consider multi-objective reinforcement learning, which arises in many real-world problems with multiple optimization goals. We approach the problem with a max-min framework focusing on fairness among the multiple goals and develop a relevant theory and a practical model-free algorithm under the max-min framework. The developed theory provides a theoretical advance in multi-objective reinforcement learning, and the proposed algorithm demonstrates a notable performance improvement over existing baseline methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-park24b, title = {The Max-Min Formulation of Multi-Objective Reinforcement Learning: From Theory to a Model-Free Algorithm}, author = {Park, Giseung and Byeon, Woohyeon and Kim, Seongmin and Havakuk, Elad and Leshem, Amir and Sung, Youngchul}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {39616--39642}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/park24b/park24b.pdf}, url = {https://proceedings.mlr.press/v235/park24b.html}, abstract = {In this paper, we consider multi-objective reinforcement learning, which arises in many real-world problems with multiple optimization goals. We approach the problem with a max-min framework focusing on fairness among the multiple goals and develop a relevant theory and a practical model-free algorithm under the max-min framework. The developed theory provides a theoretical advance in multi-objective reinforcement learning, and the proposed algorithm demonstrates a notable performance improvement over existing baseline methods.} }
Endnote
%0 Conference Paper %T The Max-Min Formulation of Multi-Objective Reinforcement Learning: From Theory to a Model-Free Algorithm %A Giseung Park %A Woohyeon Byeon %A Seongmin Kim %A Elad Havakuk %A Amir Leshem %A Youngchul Sung %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-park24b %I PMLR %P 39616--39642 %U https://proceedings.mlr.press/v235/park24b.html %V 235 %X In this paper, we consider multi-objective reinforcement learning, which arises in many real-world problems with multiple optimization goals. We approach the problem with a max-min framework focusing on fairness among the multiple goals and develop a relevant theory and a practical model-free algorithm under the max-min framework. The developed theory provides a theoretical advance in multi-objective reinforcement learning, and the proposed algorithm demonstrates a notable performance improvement over existing baseline methods.
APA
Park, G., Byeon, W., Kim, S., Havakuk, E., Leshem, A. & Sung, Y.. (2024). The Max-Min Formulation of Multi-Objective Reinforcement Learning: From Theory to a Model-Free Algorithm. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:39616-39642 Available from https://proceedings.mlr.press/v235/park24b.html.

Related Material