BriefGPT.xyz
Feb, 2021
ASAM：适应性锐度感知极小化方法用于深度神经网络的尺度不变学习
ASAM: Adaptive Sharpness-Aware Minimization for Scale-Invariant Learning of Deep Neural Networks
HTML
PDF
Jungmin Kwon, Jeongseop Kim, Hyunseo Park, In Kwon Choi
TL;DR
本论文提出了自适应锐度的概念和相应的泛化界限，并提出了利用该泛化界限的新型学习方法ASAM。在各种基准数据集上的实验证明，ASAM显著提高了模型的泛化性能。
Abstract
Recently,
learning algorithms
motivated from
sharpness
of loss surface as an effective measure of
generalization gap
have shown state-of-t
→