Guardado en:
Detalles Bibliográficos
Autores principales: Ren, Xiaoxing, Bastianello, Nicola, Parisini, Thomas, Malikopoulos, Andreas A.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2510.19199
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866909863377895424
author Ren, Xiaoxing
Bastianello, Nicola
Parisini, Thomas
Malikopoulos, Andreas A.
author_facet Ren, Xiaoxing
Bastianello, Nicola
Parisini, Thomas
Malikopoulos, Andreas A.
contents In this paper, we study the problem of reinforcement learning in multi-agent systems where communication among agents is limited. We develop a decentralized actor-critic learning framework in which each agent performs several local updates of its policy and value function, where the latter is approximated by a multi-layer neural network, before exchanging information with its neighbors. This local training strategy substantially reduces the communication burden while maintaining coordination across the network. We establish finite-time convergence analysis for the algorithm under Markov-sampling. Specifically, to attain the $\varepsilon$-accurate stationary point, the sample complexity is of order $\mathcal{O}(\varepsilon^{-3})$ and the communication complexity is of order $\mathcal{O}(\varepsilon^{-1}τ^{-1})$, where tau denotes the number of local training steps. We also show how the final error bound depends on the neural network's approximation quality. Numerical experiments in a cooperative control setting illustrate and validate the theoretical findings.
format Preprint
id arxiv_https___arxiv_org_abs_2510_19199
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Communication-Efficient Decentralized Actor-Critic Algorithm
Ren, Xiaoxing
Bastianello, Nicola
Parisini, Thomas
Malikopoulos, Andreas A.
Machine Learning
Optimization and Control
In this paper, we study the problem of reinforcement learning in multi-agent systems where communication among agents is limited. We develop a decentralized actor-critic learning framework in which each agent performs several local updates of its policy and value function, where the latter is approximated by a multi-layer neural network, before exchanging information with its neighbors. This local training strategy substantially reduces the communication burden while maintaining coordination across the network. We establish finite-time convergence analysis for the algorithm under Markov-sampling. Specifically, to attain the $\varepsilon$-accurate stationary point, the sample complexity is of order $\mathcal{O}(\varepsilon^{-3})$ and the communication complexity is of order $\mathcal{O}(\varepsilon^{-1}τ^{-1})$, where tau denotes the number of local training steps. We also show how the final error bound depends on the neural network's approximation quality. Numerical experiments in a cooperative control setting illustrate and validate the theoretical findings.
title A Communication-Efficient Decentralized Actor-Critic Algorithm
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2510.19199