Variance Reduced Coordinate Descent with Acceleration: New Method With a Surprising Application to Finite-Sum Problems
Optimization and Control
2020-02-13 v1 Machine Learning
Abstract
We propose an accelerated version of stochastic variance reduced coordinate descent -- ASVRCD. As other variance reduced coordinate descent methods such as SEGA or SVRCD, our method can deal with problems that include a non-separable and non-smooth regularizer, while accessing a random block of partial derivatives in each iteration only. However, ASVRCD incorporates Nesterov's momentum, which offers favorable iteration complexity guarantees over both SEGA and SVRCD. As a by-product of our theory, we show that a variant of Allen-Zhu (2017) is a specific case of ASVRCD, recovering the optimal oracle complexity for the finite sum objective.
Cite
@article{arxiv.2002.04670,
title = {Variance Reduced Coordinate Descent with Acceleration: New Method With a Surprising Application to Finite-Sum Problems},
author = {Filip Hanzely and Dmitry Kovalev and Peter Richtarik},
journal= {arXiv preprint arXiv:2002.04670},
year = {2020}
}
Comments
30 pages, 8 figures