Saved in:
Bibliographic Details
Main Authors: Landers, Matthew, Killian, Taylor W., Barnes, Hugo, Hartvigsen, Thomas, Doryab, Afsaneh
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2410.21151
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917188640702464
author Landers, Matthew
Killian, Taylor W.
Barnes, Hugo
Hartvigsen, Thomas
Doryab, Afsaneh
author_facet Landers, Matthew
Killian, Taylor W.
Barnes, Hugo
Hartvigsen, Thomas
Doryab, Afsaneh
contents Offline reinforcement learning in high-dimensional, discrete action spaces is challenging due to the exponential scaling of the joint action space with the number of sub-actions and the complexity of modeling sub-action dependencies. Existing methods either exhaustively evaluate the action space, making them computationally infeasible, or factorize Q-values, failing to represent joint sub-action effects. We propose Branch Value Estimation (BraVE), a value-based method that uses tree-structured action traversal to evaluate a linear number of joint actions while preserving dependency structure. BraVE outperforms prior offline RL methods by up to $20\times$ in environments with over four million actions.
format Preprint
id arxiv_https___arxiv_org_abs_2410_21151
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle BraVE: Offline Reinforcement Learning for Discrete Combinatorial Action Spaces
Landers, Matthew
Killian, Taylor W.
Barnes, Hugo
Hartvigsen, Thomas
Doryab, Afsaneh
Machine Learning
Offline reinforcement learning in high-dimensional, discrete action spaces is challenging due to the exponential scaling of the joint action space with the number of sub-actions and the complexity of modeling sub-action dependencies. Existing methods either exhaustively evaluate the action space, making them computationally infeasible, or factorize Q-values, failing to represent joint sub-action effects. We propose Branch Value Estimation (BraVE), a value-based method that uses tree-structured action traversal to evaluate a linear number of joint actions while preserving dependency structure. BraVE outperforms prior offline RL methods by up to $20\times$ in environments with over four million actions.
title BraVE: Offline Reinforcement Learning for Discrete Combinatorial Action Spaces
topic Machine Learning
url https://arxiv.org/abs/2410.21151