Study Establishes Sharp Risk Bounds for Tail-Averaged SGD in Linear Regression
A new paper analyzes the excess risk of constant-stepsize online SGD with tail-averaging for high-dimensional linear regression, providing sharp bounds that address the burn-in gap.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-11 05:00 · arXiv stat.ML
Settling the Burn-in Gap: Sharp Risk Bounds for Tail-Averaged SGD in Linear Regression