GLDAP: Global Dynamic Action Persistence Adaptation for Deep Reinforcement Learning
ACM Transactions on Autonomous and Adaptive Systems · 30.06.2023 · 18 (2) · 1-18
DOI: 10.1145/3590154 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Autonomous and Adaptive Systems, online erschienen am 28.05.2023. Verfasst von Junbo Tong und 3 weiteren Autor:innen. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 36 Literaturangaben. Gefördert von: National Key Research & Development Program of China.
Abstract (Originalsprache)
In the implementation of deep reinforcement learning (DRL), action persistence strategies are often adopted so agents maintain their actions for a fixed or variable number of steps. The choice of the persistent duration for agent actions usually has notable effects on the performance of reinforcement learning algorithms. Aiming at the research gap of global dynamic optimal action persistence and its application in multi-agent systems, we propose a novel framework: global dynamic action persistence (GLDAP), which achieves global action persistence adaptation for deep reinforcement learning. We introduce a closed-loop method that is used to learn the estimated value and the corresponding policy of each candidate action persistence. Our experiment shows that GLDAP achieves an average of 2.5%~90.7% performance improvement and 3~20 times higher sampling efficiency over several baselines across various single-agent and multi-agent domains. We also validate the ability of GLDAP to determine the optimal action persistence through multiple experiments.
Förderung
- National Key Research & Development Program of China
Zitieren
APA
Tong, J., Shi, D., Liu, Y., & Fan, W. (2023). GLDAP: Global Dynamic Action Persistence Adaptation for Deep Reinforcement Learning. ACM Transactions on Autonomous and Adaptive Systems, 18(2), 1-18. https://doi.org/10.1145/3590154
Vancouver
Tong J, Shi D, Liu Y, Fan W. GLDAP: Global Dynamic Action Persistence Adaptation for Deep Reinforcement Learning. ACM Transactions on Autonomous and Adaptive Systems. 2023;18(2):1-18. doi:10.1145/3590154
BibTeX
@article{Tong2023,
author = {Tong, Junbo and Shi, Daming and Liu, Yi and Fan, Wenhui},
title = {GLDAP: Global Dynamic Action Persistence Adaptation for Deep Reinforcement Learning},
journal = {ACM Transactions on Autonomous and Adaptive Systems},
year = {2023},
volume = {18},
number = {2},
pages = {1--18},
doi = {10.1145/3590154}
}
EUVIMED