From understanding the spreading of an epidemic to optimizing traffic flow or biological swarming, multi-agent systems are ubiquitous in nature and engineering. However, if recent progress in artificial intelligence and in particular reinforcement learning has allowed to solve complex games such as Go, Starcraft and Poker, recent methods still struggle to tackle applications with more than a dozen a players. This difficulty is known as the curse of many agents: when the number of agents increases, the game becomes computationally way harder to solve as interactions among players become intractable.In this Ph.D. thesis, we study how reinforcement learning and mean field games can benefit mutually from each other. On one hand, Mean Field Game...
Mean-Field Game (MFG) serves as a crucial mathematical framework in modeling the collective behavior...
Multi-agent reinforcement learning (RL) solves the problem of how each agent should behave optimally...
The recent mean field game (MFG) formalism facilitates otherwise intractable computation of approxim...
Mean Field Games (MFG) are a class of differential games in which each agent is infinitesimal and in...
Mean Field Games (MFGs) have been introduced to efficiently approximate games with very large popula...
Les jeux à champ moyen (MFG) sont une classe de jeux différentiels dans lequel chaque agent est infi...
Existing multi-agent reinforcement learning methods are limited typically to a small number of agent...
We address scaling up equilibrium computation in Mean Field Games (MFGs) using Online Mirror Descent...
Mean Field Games (MFGs) can potentially scale multi-agent systems to extremely large populations of ...
Learning by experience in Multi-Agent Systems (MAS) is a difficult and exciting task, due to the lac...
International audienceWe present a method enabling a large number of agents to learn how to flock, w...
openThe purpose of this paper is to develop the mean field game theory with several populations. In ...
In this manuscript, we develop reinforcement learning theory and algorithms for differential games w...
This thesis focuses on Mean Field Game (MFG) theory with applications to consensus, flocking, leader...
Mean Field Game (MFG) systems describe equilibrium configurations in differential games with infinit...
Mean-Field Game (MFG) serves as a crucial mathematical framework in modeling the collective behavior...
Multi-agent reinforcement learning (RL) solves the problem of how each agent should behave optimally...
The recent mean field game (MFG) formalism facilitates otherwise intractable computation of approxim...
Mean Field Games (MFG) are a class of differential games in which each agent is infinitesimal and in...
Mean Field Games (MFGs) have been introduced to efficiently approximate games with very large popula...
Les jeux à champ moyen (MFG) sont une classe de jeux différentiels dans lequel chaque agent est infi...
Existing multi-agent reinforcement learning methods are limited typically to a small number of agent...
We address scaling up equilibrium computation in Mean Field Games (MFGs) using Online Mirror Descent...
Mean Field Games (MFGs) can potentially scale multi-agent systems to extremely large populations of ...
Learning by experience in Multi-Agent Systems (MAS) is a difficult and exciting task, due to the lac...
International audienceWe present a method enabling a large number of agents to learn how to flock, w...
openThe purpose of this paper is to develop the mean field game theory with several populations. In ...
In this manuscript, we develop reinforcement learning theory and algorithms for differential games w...
This thesis focuses on Mean Field Game (MFG) theory with applications to consensus, flocking, leader...
Mean Field Game (MFG) systems describe equilibrium configurations in differential games with infinit...
Mean-Field Game (MFG) serves as a crucial mathematical framework in modeling the collective behavior...
Multi-agent reinforcement learning (RL) solves the problem of how each agent should behave optimally...
The recent mean field game (MFG) formalism facilitates otherwise intractable computation of approxim...