Id: arxiv/2611.19887 Url: https://arxiv.org/abs/2611.19887 Tags: math.OC, cs.LG Score: 0 Title: First-order smoothed exact-penalty methods for equality-constrained regularized nonconvex optimization Domain: math.OC Source: arxiv Authors: Shiji Zuo, Qihang Lin, Wei Liu, Yangyang Xu Reviews: 0 Subject: math.OC Category: research Comments: 0 Disputed: false Division: math Mentions: 0 Discussed: 0 Published: 2026-11-15 Page: https://delphi.science/r/arxiv/2611.19887 We consider regularized nonconvex optimization over a closed convex set $\mathcal{X} \subseteq \mathbb{R}^d$, which involves a convex nonsmooth regularizer $r$ in the objective and $n$ smooth equality constraints. To find a near-KKT point of the problem, we propose a prox-linear first-order method, named as SEPi-Prox, based on a smoothed exact-penalty formulation. At each outer iteration of SEPi-Prox, a strongly convex prox-linear subproblem is formed and solved inexactly to a sufficient accuracy. Under a local non-singularity condition that is weaker than some other commonly used regularity…