Bridging Mean-Field Games and Normalizing Flows with Trajectory Regularization

TL;DR

This paper establishes a novel connection between mean-field games and normalizing flows, enabling high-dimensional problem solving and improved model regularization through trajectory-based formulations and transport cost regularization.

Contribution

It reformulates MFGs as a trajectory-based problem parameterized by flow architectures, bridging the two models and enhancing high-dimensional density modeling and generalization.

Findings

Expressive NFs solve high-dimensional MFGs effectively.

Trajectory-based formulation improves population dynamics approximation.

Transport cost regularization enhances model generalization.

Abstract

Mean-field games (MFGs) are a modeling framework for systems with a large number of interacting agents. They have applications in economics, finance, and game theory. Normalizing flows (NFs) are a family of deep generative models that compute data likelihoods by using an invertible mapping, which is typically parameterized by using neural networks. They are useful for density modeling and data generation. While active research has been conducted on both models, few noted the relationship between the two. In this work, we unravel the connections between MFGs and NFs by contextualizing the training of an NF as solving the MFG. This is achieved by reformulating the MFG problem in terms of agent trajectories and parameterizing a discretization of the resulting MFG with flow architectures. With this connection, we explore two research directions. First, we employ expressive NF architectures to accurately solve high-dimensional MFGs, sidestepping the curse of dimensionality in traditional numerical methods. Compared with other deep learning approaches, our trajectory-based formulation encodes the continuity equation in the neural network, resulting in a better approximation of the population dynamics. Second, we regularize the training of NFs with transport costs and show the effectiveness on controlling the model's Lipschitz bound, resulting in better generalization performance. We demonstrate numerical results through comprehensive experiments on a variety of synthetic and real-life datasets.