LEAD: Min-Max Optimization from a Physical Perspective
From MaRDI portal
Publication:6352287
arXiv2010.13846MaRDI QIDQ6352287
Author name not available (Why is that?)
Publication date: 26 October 2020
Abstract: Adversarial formulations such as generative adversarial networks (GANs) have rekindled interest in two-player min-max games. A central obstacle in the optimization of such games is the rotational dynamics that hinder their convergence. In this paper, we show that game optimization shares dynamic properties with particle systems subject to multiple forces, and one can leverage tools from physics to improve optimization dynamics. Inspired by the physical framework, we propose LEAD, an optimizer for min-max games. Next, using Lyapunov stability theory and spectral analysis, we study LEAD's convergence properties in continuous and discrete time settings for a class of quadratic min-max games to demonstrate linear convergence to the Nash equilibrium. Finally, we empirically evaluate our method on synthetic setups and CIFAR-10 image generation to demonstrate improvements in GAN training.
Has companion code repository: https://github.com/ReyhaneAskari/Least_action_dynamics_minmax
This page was built for publication: LEAD: Min-Max Optimization from a Physical Perspective
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6352287)
