BriefGPT.xyz
Mar, 2024
通过随机递归方程分析随机梯度下降的重尾特性
Analysing heavy-tail properties of Stochastic Gradient Descent by means of Stochastic Recurrence Equations
HTML
PDF
Ewa Damek, Sebastian Mentemeier
TL;DR
在这篇论文中，我们回答了引用论文中的几个未解决问题，并应用不可约-近似(i-p)矩阵的理论来扩展他们的结果。
Abstract
In recent works on the theory of
machine learning
, it has been observed that heavy tail properties of
stochastic gradient descent
(SGD) can be studied in the probabilistic framework of
→