A preconditioned second-order convex splitting algorithm with extrapolation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Shen, Xinhua, Sun, Hongpeng
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915683480109056
author Shen, Xinhua
Sun, Hongpeng
author_facet Shen, Xinhua
Sun, Hongpeng
contents Nonconvex optimization problems are widespread in modern machine learning and data science. We introduce an extrapolation strategy into a class of preconditioned second-order convex splitting algorithms for nonconvex optimization problems. The proposed algorithms combine second-order backward differentiation formulas (BDF2) with an extrapolation method. Meanwhile, the implicit-explicit scheme simplifies the subproblem through a preconditioned process. As a result, our approach solves nonconvex problems efficiently without significant computational overhead. Theoretical analysis establishes global convergence of the algorithms using Kurdyka-Łojasiewicz properties. Numerical experiments include a benchmark problem, the least squares problem with SCAD regularization, and an image segmentation problem. These results demonstrate that our algorithms are highly efficient, as they achieve reduced solution times and competitive performance.
format Preprint
id arxiv_https___arxiv_org_abs_2512_14468
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A preconditioned second-order convex splitting algorithm with extrapolation
Shen, Xinhua
Sun, Hongpeng
Optimization and Control
Numerical Analysis
Nonconvex optimization problems are widespread in modern machine learning and data science. We introduce an extrapolation strategy into a class of preconditioned second-order convex splitting algorithms for nonconvex optimization problems. The proposed algorithms combine second-order backward differentiation formulas (BDF2) with an extrapolation method. Meanwhile, the implicit-explicit scheme simplifies the subproblem through a preconditioned process. As a result, our approach solves nonconvex problems efficiently without significant computational overhead. Theoretical analysis establishes global convergence of the algorithms using Kurdyka-Łojasiewicz properties. Numerical experiments include a benchmark problem, the least squares problem with SCAD regularization, and an image segmentation problem. These results demonstrate that our algorithms are highly efficient, as they achieve reduced solution times and competitive performance.
title A preconditioned second-order convex splitting algorithm with extrapolation
topic Optimization and Control
Numerical Analysis
url https://arxiv.org/abs/2512.14468