Coordinate projected gradient descent minimization and its application to orthogonal nonnegative matrix factorization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chorobura, Flavia, Lupu, Daniela, Necoara, Ion
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912303744548864
author Chorobura, Flavia
Lupu, Daniela
Necoara, Ion
author_facet Chorobura, Flavia
Lupu, Daniela
Necoara, Ion
contents In this paper we consider large-scale composite nonconvex optimization problems having the objective function formed as a sum of three terms, first has block coordinate-wise Lipschitz continuous gradient, second is twice differentiable but nonseparable and third is the indicator function of some separable closed convex set. Under these general settings we derive and analyze a new cyclic coordinate descent method, which uses the partial gradient of the differentiable part of the objective, yielding a coordinate gradient descent scheme with a novel adaptive stepsize rule. We prove that this stepsize rule makes the coordinate gradient scheme a descent method, provided that additional assumptions hold for the second term in the objective function. We also present a worst-case complexity analysis for this new method in the nonconvex settings. Numerical results on orthogonal nonnegative matrix factorization problem also confirm the efficiency of our algorithm.
format Preprint
id arxiv_https___arxiv_org_abs_2504_00770
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Coordinate projected gradient descent minimization and its application to orthogonal nonnegative matrix factorization
Chorobura, Flavia
Lupu, Daniela
Necoara, Ion
Optimization and Control
In this paper we consider large-scale composite nonconvex optimization problems having the objective function formed as a sum of three terms, first has block coordinate-wise Lipschitz continuous gradient, second is twice differentiable but nonseparable and third is the indicator function of some separable closed convex set. Under these general settings we derive and analyze a new cyclic coordinate descent method, which uses the partial gradient of the differentiable part of the objective, yielding a coordinate gradient descent scheme with a novel adaptive stepsize rule. We prove that this stepsize rule makes the coordinate gradient scheme a descent method, provided that additional assumptions hold for the second term in the objective function. We also present a worst-case complexity analysis for this new method in the nonconvex settings. Numerical results on orthogonal nonnegative matrix factorization problem also confirm the efficiency of our algorithm.
title Coordinate projected gradient descent minimization and its application to orthogonal nonnegative matrix factorization
topic Optimization and Control
url https://arxiv.org/abs/2504.00770