BriefGPT.xyz
Dec, 2014
深度学习中隐式正则化的作用：寻找真正的归纳偏差
In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning
HTML
PDF
Behnam Neyshabur, Ryota Tomioka, Nathan Srebro
TL;DR
文章提出了实验结果，证明其他形式的容量控制与网络规模不同，在学习多层前馈网络中起着中心作用。 作者通过类比矩阵分解来论证这是一种归纳偏差，有助于揭示深度学习的内在机理。
Abstract
We present experiments demonstrating that some other form of
capacity control
, different from network size, plays a central role in
learning
mult
→