BriefGPT.xyz
Oct, 2018
最小二乘法中提前停止的连续时间视角
A Continuous-Time View of Early Stopping for Least Squares Regression
HTML
PDF
Alnur Ali, J. Zico Kolter, Ryan J. Tibshirani
TL;DR
本文研究了应用于最小二乘回归问题的梯度下降迭代的统计特性,将其与岭回归的风险进行比较。研究发现,在梯度下降的整个路径上,其风险不低于岭回归的1.69倍,并在平均信号下保持相对风险边界,同时考虑了极限风险表达式和支撑数值实验。
Abstract
We study the statistical properties of the iterates generated by
gradient descent
, applied to the fundamental problem of
least squares regression
. We take a continuous-time view, i.e., consider infinitesimal step
→