Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Liao, Feng-Yi, Ding, Lijun, Zheng, Yang
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912873920331776
author Liao, Feng-Yi
Ding, Lijun
Zheng, Yang
author_facet Liao, Feng-Yi
Ding, Lijun
Zheng, Yang
contents Many practical optimization problems lack strong convexity. Fortunately, recent studies have revealed that first-order algorithms also enjoy linear convergences under various weaker regularity conditions. While the relationship among different conditions for convex and smooth functions is well-understood, it is not the case for the nonsmooth setting. In this paper, we go beyond convexity and smoothness, and clarify the connections among common regularity conditions in the class of weakly convex functions, including $\textit{strong convexity}$, $\textit{restricted secant inequality}$, $\textit{subdifferential error bound}$, $\textit{Polyak-Łojasiewicz inequality}$, and $\textit{quadratic growth}$. In addition, using these regularity conditions, we present a simple and modular proof for the linear convergence of the proximal point method (PPM) for convex and weakly convex optimization problems. The linear convergence also holds when the subproblems of PPM are solved inexactly with a proper control of inexactness.
format Preprint
id arxiv_https___arxiv_org_abs_2312_16775
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods
Liao, Feng-Yi
Ding, Lijun
Zheng, Yang
Optimization and Control
Systems and Control
Many practical optimization problems lack strong convexity. Fortunately, recent studies have revealed that first-order algorithms also enjoy linear convergences under various weaker regularity conditions. While the relationship among different conditions for convex and smooth functions is well-understood, it is not the case for the nonsmooth setting. In this paper, we go beyond convexity and smoothness, and clarify the connections among common regularity conditions in the class of weakly convex functions, including $\textit{strong convexity}$, $\textit{restricted secant inequality}$, $\textit{subdifferential error bound}$, $\textit{Polyak-Łojasiewicz inequality}$, and $\textit{quadratic growth}$. In addition, using these regularity conditions, we present a simple and modular proof for the linear convergence of the proximal point method (PPM) for convex and weakly convex optimization problems. The linear convergence also holds when the subproblems of PPM are solved inexactly with a proper control of inexactness.
title Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods
topic Optimization and Control
Systems and Control
url https://arxiv.org/abs/2312.16775