Bounds for the tracking error of first-order online optimization methods

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Madden, Liam, Becker, Stephen, Dall'Anese, Emiliano
Format: Preprint
Published: 2020
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914867874627584
author Madden, Liam
Becker, Stephen
Dall'Anese, Emiliano
author_facet Madden, Liam
Becker, Stephen
Dall'Anese, Emiliano
contents This paper investigates online algorithms for smooth time-varying optimization problems, focusing first on methods with constant step-size, momentum, and extrapolation-length. Assuming strong convexity, precise results for the tracking iterate error (the limit supremum of the norm of the difference between the optimal solution and the iterates) for online gradient descent are derived. The paper then considers a general first-order framework, where a universal lower bound on the tracking iterate error is established. Furthermore, a method using "long-steps" is proposed and shown to achieve the lower bound up to a fixed constant. This method is then compared with online gradient descent for specific examples. Finally, the paper analyzes the effect of regularization when the cost is not strongly convex. With regularization, it is possible to achieve a non-regret bound. The paper ends by testing the accelerated and regularized methods on synthetic time-varying least-squares and logistic regression problems, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2003_02400
institution arXiv
publishDate 2020
record_format arxiv
spellingShingle Bounds for the tracking error of first-order online optimization methods
Madden, Liam
Becker, Stephen
Dall'Anese, Emiliano
Optimization and Control
This paper investigates online algorithms for smooth time-varying optimization problems, focusing first on methods with constant step-size, momentum, and extrapolation-length. Assuming strong convexity, precise results for the tracking iterate error (the limit supremum of the norm of the difference between the optimal solution and the iterates) for online gradient descent are derived. The paper then considers a general first-order framework, where a universal lower bound on the tracking iterate error is established. Furthermore, a method using "long-steps" is proposed and shown to achieve the lower bound up to a fixed constant. This method is then compared with online gradient descent for specific examples. Finally, the paper analyzes the effect of regularization when the cost is not strongly convex. With regularization, it is possible to achieve a non-regret bound. The paper ends by testing the accelerated and regularized methods on synthetic time-varying least-squares and logistic regression problems, respectively.
title Bounds for the tracking error of first-order online optimization methods
topic Optimization and Control
url https://arxiv.org/abs/2003.02400