Algorithms of Robust Stochastic Optimization Based on Mirror Descent Method
Statistics Theory
2019-07-08 v1 Optimization and Control
Statistics Theory
Abstract
We propose an approach to construction of robust non-Euclidean iterative algorithms for convex composite stochastic optimization based on truncation of stochastic gradients. For such algorithms, we establish sub-Gaussian confidence bounds under weak assumptions about the tails of the noise distribution in convex and strongly convex settings. Robust estimates of the accuracy of general stochastic algorithms are also proposed.
Cite
@article{arxiv.1907.02707,
title = {Algorithms of Robust Stochastic Optimization Based on Mirror Descent Method},
author = {Anatoli Juditsky and Alexander Nazin and Arkadi Nemirovsky and Alexandre Tsybakov},
journal= {arXiv preprint arXiv:1907.02707},
year = {2019}
}
Comments
Automation and Remote Control / Avtomatika i Telemekhanika, MAIK Nauka/Interperiodica, In press