English

Finite state N-agent and mean field control problems

Optimization and Control 2021-03-29 v2 Analysis of PDEs Probability

Abstract

We examine mean field control problems on a finite state space, in continuous time and over a finite time horizon. We characterize the value function of the mean field control problem as the unique viscosity solution of a Hamilton-Jacobi-Bellman equation in the simplex. In absence of any convexity assumption, we exploit this characterization to prove convergence, as NN grows, of the value functions of the centralized NN-agent optimal control problem to the limit mean field control problem value function, with a convergence rate of order 1/N1/\sqrt{N}. Then, assuming convexity, we show that the limit value function is smooth and establish propagation of chaos, i.e. convergence of the NN-agent optimal trajectory to the unique limiting optimal trajectory, with an explicit rate.

Keywords

Cite

@article{arxiv.2010.11569,
  title  = {Finite state N-agent and mean field control problems},
  author = {Alekos Cecchin},
  journal= {arXiv preprint arXiv:2010.11569},
  year   = {2021}
}
R2 v1 2026-06-23T19:32:55.296Z