Constrained Episodic Reinforcement Learning in Concave-Convex and Knapsack Settings
Abstract
We propose an algorithm for tabular episodic reinforcement learning with constraints. We provide a modular analysis with strong theoretical guarantees for settings with concave rewards and convex constraints, and for settings with hard constraints (knapsacks). Most of the previous work in constrained reinforcement learning is limited to linear constraints, and the remaining work focuses on either the feasibility question or settings with a single episode. Our experiments demonstrate that the proposed algorithm significantly outperforms these approaches in existing constrained episodic environments.
Cite
Text
Brantley et al. "Constrained Episodic Reinforcement Learning in Concave-Convex and Knapsack Settings." Neural Information Processing Systems, 2020.Markdown
[Brantley et al. "Constrained Episodic Reinforcement Learning in Concave-Convex and Knapsack Settings." Neural Information Processing Systems, 2020.](https://mlanthology.org/neurips/2020/brantley2020neurips-constrained/)BibTeX
@inproceedings{brantley2020neurips-constrained,
title = {{Constrained Episodic Reinforcement Learning in Concave-Convex and Knapsack Settings}},
author = {Brantley, Kianté and Dudik, Miro and Lykouris, Thodoris and Miryoosefi, Sobhan and Simchowitz, Max and Slivkins, Aleksandrs and Sun, Wen},
booktitle = {Neural Information Processing Systems},
year = {2020},
url = {https://mlanthology.org/neurips/2020/brantley2020neurips-constrained/}
}