Computer Science > Machine Learning
[Submitted on 6 Feb 2021 (v1), last revised 8 Jun 2023 (this version, v19)]
Title:Rethinking the Implementation Tricks and Monotonicity Constraint in Cooperative Multi-Agent Reinforcement Learning
View PDFAbstract:Many complex multi-agent systems such as robot swarms control and autonomous vehicle coordination can be modeled as Multi-Agent Reinforcement Learning (MARL) tasks. QMIX, a widely popular MARL algorithm, has been used as a baseline for the benchmark environments, e.g., Starcraft Multi-Agent Challenge (SMAC), Difficulty-Enhanced Predator-Prey (DEPP). Recent variants of QMIX target relaxing the monotonicity constraint of QMIX, allowing for performance improvement in SMAC. In this paper, we investigate the code-level optimizations of these variants and the monotonicity constraint. (1) We find that such improvements of the variants are significantly affected by various code-level optimizations. (2) The experiment results show that QMIX with normalized optimizations outperforms other works in SMAC; (3) beyond the common wisdom from these works, the monotonicity constraint can improve sample efficiency in SMAC and DEPP. We also discuss why monotonicity constraints work well in purely cooperative tasks with a theoretical analysis. We open-source the code at \url{this https URL}.
Submission history
From: Jian Hu [view email][v1] Sat, 6 Feb 2021 02:28:09 UTC (6,958 KB)
[v2] Mon, 22 Feb 2021 14:01:38 UTC (7,756 KB)
[v3] Mon, 1 Mar 2021 15:37:05 UTC (6,161 KB)
[v4] Tue, 9 Mar 2021 13:46:31 UTC (6,575 KB)
[v5] Thu, 11 Mar 2021 08:29:17 UTC (6,766 KB)
[v6] Mon, 15 Mar 2021 06:21:44 UTC (5,946 KB)
[v7] Fri, 9 Apr 2021 03:39:11 UTC (5,946 KB)
[v8] Mon, 19 Apr 2021 09:00:01 UTC (5,578 KB)
[v9] Wed, 21 Apr 2021 08:38:21 UTC (5,578 KB)
[v10] Thu, 13 May 2021 13:06:58 UTC (370 KB)
[v11] Mon, 7 Jun 2021 07:54:45 UTC (434 KB)
[v12] Thu, 1 Jul 2021 07:37:31 UTC (449 KB)
[v13] Mon, 2 Aug 2021 00:09:58 UTC (484 KB)
[v14] Wed, 1 Sep 2021 10:51:34 UTC (363 KB)
[v15] Fri, 1 Oct 2021 00:48:44 UTC (359 KB)
[v16] Thu, 11 Nov 2021 13:08:03 UTC (574 KB)
[v17] Sat, 4 Dec 2021 03:45:57 UTC (585 KB)
[v18] Sat, 1 Jan 2022 02:00:09 UTC (398 KB)
[v19] Thu, 8 Jun 2023 07:43:05 UTC (398 KB)
Current browse context:
cs.LG
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.