BriefGPT.xyz
Jun, 2020
针对复合凸光滑优化的随机梯度方法统一分析
Unified Analysis of Stochastic Gradient Methods for Composite Convex and Smooth Optimization
HTML
PDF
Ahmed Khaled, Othmane Sebbouh, Nicolas Loizou, Robert M. Gower, Peter Richtárik
TL;DR
本文为最小化平滑和凸损失加上凸正则化的随机梯度算法提供了一致的收敛性分析定理,并探讨了特定算法的最优小批量大小。
Abstract
We present a unified theorem for the
convergence analysis
of stochastic gradient algorithms for minimizing a smooth and
convex loss
plus a
convex
→