Shadow splitting methods for nonconvex optimisation: epi-approximation, convergence and saddle point avoidance

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Atenas, Felipe
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908730520502272
author Atenas, Felipe
author_facet Atenas, Felipe
contents We propose the shadow Davis-Yin three-operator splitting method to solve nonconvex optimisation problems. Its convergence analysis is based on a merit function resembling the Moreau envelope. We explore variational analysis properties behind the merit function and the iteration operators associated with the shadow method. By capitalising on these results, we establish convergence of a damped version of the shadow method via sufficient descent of the merit function, and obtain (almost surely) guarantees of avoidance of strict saddle points of weakly convex semialgebraic optimisation problems. We perform numerical experiments for a nonconvex variable selection problem to test our findings.
format Preprint
id arxiv_https___arxiv_org_abs_2512_20433
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Shadow splitting methods for nonconvex optimisation: epi-approximation, convergence and saddle point avoidance
Atenas, Felipe
Optimization and Control
49J52, 65K05, 65K10, 90C26, 90C30
We propose the shadow Davis-Yin three-operator splitting method to solve nonconvex optimisation problems. Its convergence analysis is based on a merit function resembling the Moreau envelope. We explore variational analysis properties behind the merit function and the iteration operators associated with the shadow method. By capitalising on these results, we establish convergence of a damped version of the shadow method via sufficient descent of the merit function, and obtain (almost surely) guarantees of avoidance of strict saddle points of weakly convex semialgebraic optimisation problems. We perform numerical experiments for a nonconvex variable selection problem to test our findings.
title Shadow splitting methods for nonconvex optimisation: epi-approximation, convergence and saddle point avoidance
topic Optimization and Control
49J52, 65K05, 65K10, 90C26, 90C30
url https://arxiv.org/abs/2512.20433