Approximate Bregman Proximal Gradient Algorithm for Relatively Smooth Nonconvex Optimization

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Takahashi, Shota, Takeda, Akiko
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913582945402880
author Takahashi, Shota
Takeda, Akiko
author_facet Takahashi, Shota
Takeda, Akiko
contents In this paper, we propose the approximate Bregman proximal gradient algorithm (ABPG) for solving composite nonconvex optimization problems. ABPG employs a new distance that approximates the Bregman distance, making the subproblem of ABPG simpler to solve compared to existing Bregman-type algorithms. The subproblem of ABPG is often expressed in a closed form. Similarly to existing Bregman-type algorithms, ABPG does not require the global Lipschitz continuity for the gradient of the smooth part. Instead, assuming the smooth adaptable property, we establish the global subsequential convergence under standard assumptions. Additionally, assuming that the Kurdyka--Łojasiewicz property holds, we prove the global convergence for a special case. Our numerical experiments on the $\ell_p$ regularized least squares problem, the $\ell_p$ loss problem, and the nonnegative linear system show that ABPG outperforms existing algorithms especially when the gradient of the smooth part is not globally Lipschitz or even local Lipschitz continuous.
format Preprint
id arxiv_https___arxiv_org_abs_2311_07847
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Approximate Bregman Proximal Gradient Algorithm for Relatively Smooth Nonconvex Optimization
Takahashi, Shota
Takeda, Akiko
Optimization and Control
90C26, 49M37, 65K05
In this paper, we propose the approximate Bregman proximal gradient algorithm (ABPG) for solving composite nonconvex optimization problems. ABPG employs a new distance that approximates the Bregman distance, making the subproblem of ABPG simpler to solve compared to existing Bregman-type algorithms. The subproblem of ABPG is often expressed in a closed form. Similarly to existing Bregman-type algorithms, ABPG does not require the global Lipschitz continuity for the gradient of the smooth part. Instead, assuming the smooth adaptable property, we establish the global subsequential convergence under standard assumptions. Additionally, assuming that the Kurdyka--Łojasiewicz property holds, we prove the global convergence for a special case. Our numerical experiments on the $\ell_p$ regularized least squares problem, the $\ell_p$ loss problem, and the nonnegative linear system show that ABPG outperforms existing algorithms especially when the gradient of the smooth part is not globally Lipschitz or even local Lipschitz continuous.
title Approximate Bregman Proximal Gradient Algorithm for Relatively Smooth Nonconvex Optimization
topic Optimization and Control
90C26, 49M37, 65K05
url https://arxiv.org/abs/2311.07847