BriefGPT.xyz
Aug, 2019
通过平移Rademacher过程的快速PAC-Bayes泛化界限
Fast-rate PAC-Bayes Generalization Bounds via Shifted Rademacher Processes
HTML
PDF
Jun Yang, Shengyang Sun, Daniel M. Roy
TL;DR
本文旨在扩展Rademacher复杂性和最新PAC-贝叶斯理论之间的桥梁，首先通过平移Rademacher过程来匹配Catoni PAC-Bayes界限的快速率，然后最新地导出了快速PAC-Bayes界限，重点是后验集中在的经验风险表面的“平坦度”。
Abstract
The developments of
rademacher complexity
and
pac-bayesian theory
have been largely independent. One exception is the PAC-Bayes theorem of Kakade, Sridharan, and Tewari (2008), which is established via
→