Learning in Herding Mean Field Games: Single-Loop Algorithm with Finite-Time Convergence Analysis

TL;DR

This paper introduces a novel single-loop algorithm, ASAC-MFG, for solving a broad class of mean field games with multiple equilibria, providing finite-time convergence guarantees under minimal assumptions.

Contribution

The work expands the class of solvable mean field games to include those with multiple equilibria and introduces a provably convergent single-loop policy optimization algorithm.

Findings

ASAC-MFG converges to a mean field equilibrium within finite time.

The algorithm is applicable to average-reward MDPs without contraction assumptions.

Theoretical convergence rates match state-of-the-art for MDPs.

Abstract

We consider discrete-time stationary mean field games (MFG) with unknown dynamics and design algorithms for finding the equilibrium with finite-time complexity guarantees. Prior solutions to the problem assume either the contraction of a mean field optimality-consistency operator or strict weak monotonicity, which may be overly restrictive. In this work, we introduce a new class of solvable MFGs, named the "fully herding class", which expands the known solvable class of MFGs and for the first time includes problems with multiple equilibria. We propose a direct policy optimization method, Accelerated Single-loop Actor Critic Algorithm for Mean Field Games (ASAC-MFG), that provably finds a global equilibrium for MFGs within this class, under suitable access to a single trajectory of Markovian samples. Different from the prior methods, ASAC-MFG is single-loop and single-sample-path. We establish the finite-time and finite-sample convergence of ASAC-MFG to a mean field equilibrium via new techniques that we develop for multi-time-scale stochastic approximation. We support the theoretical results with illustrative numerical simulations. When the mean field does not affect the transition and reward, a MFG reduces to a Markov decision process (MDP) and ASAC-MFG becomes an actor-critic algorithm for finding the optimal policy in average-reward MDPs, with a sample complexity matching the state-of-the-art. Previous works derive the complexity assuming a contraction on the Bellman operator, which is invalid for average-reward MDPs. We match the rate while removing the untenable assumption through an improved Lyapunov function.