New arXiv paper presents bias-reduced, parallel heavy-tailed stochastic convex optimization in ℓp geometries
A new arXiv paper (2611.21844) studies parallel stochastic convex optimization, providing bias-reduced methods for heavy-tailed noise in ℓp geometries.
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-11-18 05:00 · arXiv stat.ML
Bias-Reduced and Parallel Heavy-Tailed Stochastic Convex Optimization in $\ell_p$ Geometries