Convergence of Gradient Methods on Bilinear Zero-Sum Games
From MaRDI portal
Publication:6323726
arXiv1908.05699MaRDI QIDQ6323726
Yaoliang Yu, Guojun Zhang
Publication date: 15 August 2019
Abstract: Min-max formulations have attracted great attention in the ML community due to the rise of deep generative models and adversarial methods, while understanding the dynamics of gradient algorithms for solving such formulations has remained a grand challenge. As a first step, we restrict to bilinear zero-sum games and give a systematic analysis of popular gradient updates, for both simultaneous and alternating versions. We provide exact conditions for their convergence and find the optimal parameter setup and convergence rates. In particular, our results offer formal evidence that alternating updates converge "better" than simultaneous ones.
Has companion code repository: https://github.com/Gordon-Guojun-Zhang/ICLR-2020
This page was built for publication: Convergence of Gradient Methods on Bilinear Zero-Sum Games
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6323726)
