BriefGPT.xyz
Jun, 2019
随机组合优化的通用加速框架
A Generic Acceleration Framework for Stochastic Composite Optimization
HTML
PDF
Andrei Kulunchakov, Julien Mairal
TL;DR
本文介绍了在目标函数为凸或强凸函数时获取加速一阶随机优化算法的各种机制，同时扩展了最初用于确定性目标的Catalyst方法到随机问题领域，并提供了一个新的关于处理不精确近端算子时的鲁棒性的泛化分析
Abstract
In this paper, we introduce various mechanisms to obtain accelerated first-order
stochastic optimization
algorithms when the objective function is convex or strongly convex. Specifically, we extend the
catalyst approach
→