BriefGPT.xyz
Mar, 2020
神经网络生成对抗模仿学习：全局最优性和收敛速率
Generative Adversarial Imitation Learning with Neural Networks: Global Optimality and Convergence Rate
HTML
PDF
Yufeng Zhang, Qi Cai, Zhuoran Yang, Zhaoran Wang
TL;DR
本文采用基于梯度的交替更新策略，分析了生成对抗学习在神经网络结构下的全局优化和收敛速率，证明了该方法的全局最优解和收敛性。
Abstract
generative adversarial imitation learning
(GAIL) demonstrates tremendous success in practice, especially when combined with
neural networks
. Different from reinforcement learning, GAIL learns both policy and rewa
→