Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods
Lists
Only you see this
Abstract
Our goal is to improve variance reducing stochastic methods through better control variates. We first propose a modification of SVRG which uses the Hessian to track gradients over time, rather than to recondition, increasing the correlation of the control variates and leading to faster theoretical convergence close to the optimum. We then propose accurate and computationally efficient approximations to the Hessian, both using a diagonal and a low-rank matrix. Finally, we demonstrate the effectiveness of our method on a wide range of problems.
BibTeX
@inproceedings{gower2018tracking,
author = {Robert M. Gower and Nicolas Le Roux and Francis R. Bach},
title = {{Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods}},
booktitle = {International Conference on Artificial Intelligence and Statistics},
year = {2018},
url = {https://openreview.net/forum?id=XsdBwt4hBxr},
}