BriefGPT.xyz
Oct, 2018
寻找生成对抗网络的混合纳什均衡
Finding Mixed Nash Equilibria of Generative Adversarial Networks
HTML
PDF
Ya-Ping Hsieh, Chen Liu, Volkan Cevher
TL;DR
通过无限维二人博弈的新算法框架，我们证明了混合Nash Equilibria上的收敛性率，最终通过简单取样程序实现了高效的算法，证明了其在速度和质量上的优越表现。
Abstract
We reconsider the training objective of
generative adversarial networks
(GANs) from the mixed
nash equilibria
(NE) perspective. Inspired by the classical
→