Delayed Feedback in Online Non-Convex Optimization: A Non-Stationary Approach with Applications

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lara, Felipe, Vega, Cristian
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909983523733504
author Lara, Felipe
Vega, Cristian
author_facet Lara, Felipe
Vega, Cristian
contents We study non-convex delayed-noise online optimization problems by evaluating dynamic regret in the non-stationary setting when the loss functions are quasar-convex. In particular, we consider scenarios involving quasar-convex functions either with a Lipschitz gradient or weakly smooth and, for each case, we ensure bounded dynamic regret in terms of cumulative path variation achieving sub-linear regret rates. Furthermore, we illustrate the flexibility of our framework by applying it to both theoretical settings such as zeroth-order (bandit) and also to practical applications with quadratic fractional functions. Moreover, we provide new examples of non-convex functions that are quasar-convex by proving that the class of differentiable strongly quasiconvex functions (Polyak 1966) are strongly quasar-convex on convex compact sets. Finally, several numerical experiments validate our theoretical findings, illustrating the effectiveness of our approach.
format Preprint
id arxiv_https___arxiv_org_abs_2412_14506
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Delayed Feedback in Online Non-Convex Optimization: A Non-Stationary Approach with Applications
Lara, Felipe
Vega, Cristian
Optimization and Control
90C26, 65K05, 90C56, 90C30, 68W27
We study non-convex delayed-noise online optimization problems by evaluating dynamic regret in the non-stationary setting when the loss functions are quasar-convex. In particular, we consider scenarios involving quasar-convex functions either with a Lipschitz gradient or weakly smooth and, for each case, we ensure bounded dynamic regret in terms of cumulative path variation achieving sub-linear regret rates. Furthermore, we illustrate the flexibility of our framework by applying it to both theoretical settings such as zeroth-order (bandit) and also to practical applications with quadratic fractional functions. Moreover, we provide new examples of non-convex functions that are quasar-convex by proving that the class of differentiable strongly quasiconvex functions (Polyak 1966) are strongly quasar-convex on convex compact sets. Finally, several numerical experiments validate our theoretical findings, illustrating the effectiveness of our approach.
title Delayed Feedback in Online Non-Convex Optimization: A Non-Stationary Approach with Applications
topic Optimization and Control
90C26, 65K05, 90C56, 90C30, 68W27
url https://arxiv.org/abs/2412.14506