May 24, 2024Open Access

A Fisher-Rao gradient flow for entropic mean-field min-max games

Key Points

Key points are not available for this paper at this time.

Abstract

Gradient flows play a substantial role in addressing many machine learning problems. We examine the convergence in continuous-time of a Fisher-Rao (Mean-Field Birth-Death) gradient flow in the context of solving convex-concave min-max games with entropy regularization. We propose appropriate Lyapunov functions to demonstrate convergence with explicit rates to the unique mixed Nash equilibrium.

Ask AI

Helpful

Bookmark

View Full Paper

Cite This Study

Lascu et al. (Fri,) studied this question.

synapsesocial.com/papers/68e68aacb6db6435876123b6 https://doi.org/https://doi.org/10.48550/arxiv.2405.15834

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

1Optimal Control and Potential Games in the Mean Field2024
2Fisher-Rao Gradient Flows of Linear Programs and State-Action Natural Policy Gradients2024
3Convergence of Time-Averaged Mean Field Gradient Descent Dynamics for Continuous Multi-Player Zero-Sum Games2025
4Learning Mean-Field Games through Mean-Field Actor-Critic Flow2025
5A Mean-Field Analysis of Neural Stochastic Gradient Descent-Ascent for Functional Minimax Optimization2024

Ask AI

Helpful

Bookmark

View Full Paper