Showing 1–2 of 2 results for author: Flem, L L
-
Sharp uniform-in-time propagation of chaos
Authors:
Daniel Lacker,
Luc Le Flem
Abstract:
We prove the optimal rate of quantitative propagation of chaos, uniformly in time, for interacting diffusions. Our main examples are interactions governed by convex potentials and models on the torus with small interactions. We show that the distance between the $k$-particle marginal of the $n$-particle system and its limiting product measure is $O((k/n)^2)$, uniformly in time, with distance measu…
▽ More
We prove the optimal rate of quantitative propagation of chaos, uniformly in time, for interacting diffusions. Our main examples are interactions governed by convex potentials and models on the torus with small interactions. We show that the distance between the $k$-particle marginal of the $n$-particle system and its limiting product measure is $O((k/n)^2)$, uniformly in time, with distance measured either by relative entropy, squared quadratic Wasserstein metric, or squared total variation. Our proof is based on an analysis of relative entropy through the BBGKY hierarchy, adapting prior work of the first author to the time-uniform case by means of log-Sobolev inequalities.
△ Less
Submitted 20 January, 2023; v1 submitted 24 May, 2022;
originally announced May 2022.
-
Closed-loop convergence for mean field games with common noise
Authors:
Daniel Lacker,
Luc Le Flem
Abstract:
This paper studies the convergence problem for mean field games with common noise. We define a suitable notion of weak mean field equilibria, which we prove captures all subsequential limit points, as $n\to\infty$, of closed-loop approximate equilibria from the corresponding $n$-player games. This extends to the common noise setting a recent result of the first author, while also simplifying a key…
▽ More
This paper studies the convergence problem for mean field games with common noise. We define a suitable notion of weak mean field equilibria, which we prove captures all subsequential limit points, as $n\to\infty$, of closed-loop approximate equilibria from the corresponding $n$-player games. This extends to the common noise setting a recent result of the first author, while also simplifying a key step in the proof and allowing unbounded coefficients and non-i.i.d. initial conditions. Conversely, we show that every weak mean field equilibrium arises as the limit of some sequence of approximate equilibria for the $n$-player games, as long as the latter are formulated over a broader class of closed-loop strategies which may depend on an additional common signal.
△ Less
Submitted 19 August, 2022; v1 submitted 7 July, 2021;
originally announced July 2021.
