Decentralized multi-agent reinforcement learning based on best-response policies

Introduction: Multi-agent systems are an interdisciplinary research field that describes the concept of multiple decisive individuals interacting with a usually partially observable environment. Given the recent advances in single-agent reinforcement learning, multi-agent reinforcement learning (RL)...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Frontiers in robotics and AI Jg. 11; S. 1229026
Hauptverfasser: Gabler, Volker, Wollherr, Dirk
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Switzerland Frontiers Media S.A 16.04.2024
Schlagworte:
actor–critic algorithm
deep learning, artificial intelligence
game theory
multi-agent
multi-agent reinforcement learning
Robotics and AI
Stackelberg
deep learning, artificial intelligence
multi-agent reinforcement learning
Stackelberg
multi-agent
reinforcement leaning
game theory
decentralized learning schemes
actor–critic algorithm
ISSN:2296-9144, 2296-9144
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Introduction: Multi-agent systems are an interdisciplinary research field that describes the concept of multiple decisive individuals interacting with a usually partially observable environment. Given the recent advances in single-agent reinforcement learning, multi-agent reinforcement learning (RL) has gained tremendous interest in recent years. Most research studies apply a fully centralized learning scheme to ease the transfer from the single-agent domain to multi-agent systems. Methods: In contrast, we claim that a decentralized learning scheme is preferable for applications in real-world scenarios as this allows deploying a learning algorithm on an individual robot rather than deploying the algorithm to a complete fleet of robots. Therefore, this article outlines a novel actor–critic (AC) approach tailored to cooperative MARL problems in sparsely rewarded domains. Our approach decouples the MARL problem into a set of distributed agents that model the other agents as responsive entities. In particular, we propose using two separate critics per agent to distinguish between the joint task reward and agent-based costs as commonly applied within multi-robot planning. On one hand, the agent-based critic intends to decrease agent-specific costs. On the other hand, each agent intends to optimize the joint team reward based on the joint task critic. As this critic still depends on the joint action of all agents, we outline two suitable behavior models based on Stackelberg games: a game against nature and a dyadic game against each agent. Following these behavior models, our algorithm allows fully decentralized execution and training. Results and Discussion: We evaluate our presented method using the proposed behavior models within a sparsely rewarded simulated multi-agent environment. Although our approach already outperforms the state-of-the-art learners, we conclude this article by outlining possible extensions of our algorithm that future research may build upon.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
Eiji Uchibe, Advanced Telecommunications Research Institute International (ATR), Japan
Reviewed by: Laura Ferrarotti, Bruno Kessler Foundation (FBK), Italy
Edited by: Giovanni Iacca, University of Trento, Italy
ISSN:2296-9144
2296-9144
DOI:10.3389/frobt.2024.1229026