Mean-Field Sampling for Cooperative Multi-Agent Reinforcement Learning

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Anand, Emile, Karmarkar, Ishani, Qu, Guannan
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915572587954176
author Anand, Emile
Karmarkar, Ishani
Qu, Guannan
author_facet Anand, Emile
Karmarkar, Ishani
Qu, Guannan
contents Designing efficient algorithms for multi-agent reinforcement learning (MARL) is fundamentally challenging because the size of the joint state and action spaces grows exponentially in the number of agents. These difficulties are exacerbated when balancing sequential global decision-making with local agent interactions. In this work, we propose a new algorithm $\texttt{SUBSAMPLE-MFQ}$ ($\textbf{Subsample}$-$\textbf{M}$ean-$\textbf{F}$ield-$\textbf{Q}$-learning) and a decentralized randomized policy for a system with $n$ agents. For any $k\leq n$, our algorithm learns a policy for the system in time polynomial in $k$. We prove that this learned policy converges to the optimal policy on the order of $\tilde{O}(1/\sqrt{k})$ as the number of subsampled agents $k$ increases. In particular, this bound is independent of the number of agents $n$.
format Preprint
id arxiv_https___arxiv_org_abs_2412_00661
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Mean-Field Sampling for Cooperative Multi-Agent Reinforcement Learning
Anand, Emile
Karmarkar, Ishani
Qu, Guannan
Machine Learning
Artificial Intelligence
Multiagent Systems
Systems and Control
Optimization and Control
60J20, 68T99
I.2.11
Designing efficient algorithms for multi-agent reinforcement learning (MARL) is fundamentally challenging because the size of the joint state and action spaces grows exponentially in the number of agents. These difficulties are exacerbated when balancing sequential global decision-making with local agent interactions. In this work, we propose a new algorithm $\texttt{SUBSAMPLE-MFQ}$ ($\textbf{Subsample}$-$\textbf{M}$ean-$\textbf{F}$ield-$\textbf{Q}$-learning) and a decentralized randomized policy for a system with $n$ agents. For any $k\leq n$, our algorithm learns a policy for the system in time polynomial in $k$. We prove that this learned policy converges to the optimal policy on the order of $\tilde{O}(1/\sqrt{k})$ as the number of subsampled agents $k$ increases. In particular, this bound is independent of the number of agents $n$.
title Mean-Field Sampling for Cooperative Multi-Agent Reinforcement Learning
topic Machine Learning
Artificial Intelligence
Multiagent Systems
Systems and Control
Optimization and Control
60J20, 68T99
I.2.11
url https://arxiv.org/abs/2412.00661