KKT-based optimality conditions for neural network approximation

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Peiris, Vinesha, Sukhorukova, Nadezda, Ugon, Julien
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866911016247361536
author Peiris, Vinesha
Sukhorukova, Nadezda
Ugon, Julien
author_facet Peiris, Vinesha
Sukhorukova, Nadezda
Ugon, Julien
contents In this paper, we obtain necessary optimality conditions for neural network approximation. We consider neural networks in Manhattan ($l_1$ norm) and Chebyshev ($\max$ norm). The optimality conditions are based on neural networks with at most one hidden layer. We reformulate nonsmooth unconstrained optimisation problems as larger dimension constrained problems with smooth objective functions and constraints. Then we use KKT conditions to develop the necessary conditions and present the optimality conditions in terms of convex analysis and convex sets.
format Preprint
id arxiv_https___arxiv_org_abs_2506_17305
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle KKT-based optimality conditions for neural network approximation
Peiris, Vinesha
Sukhorukova, Nadezda
Ugon, Julien
Optimization and Control
In this paper, we obtain necessary optimality conditions for neural network approximation. We consider neural networks in Manhattan ($l_1$ norm) and Chebyshev ($\max$ norm). The optimality conditions are based on neural networks with at most one hidden layer. We reformulate nonsmooth unconstrained optimisation problems as larger dimension constrained problems with smooth objective functions and constraints. Then we use KKT conditions to develop the necessary conditions and present the optimality conditions in terms of convex analysis and convex sets.
title KKT-based optimality conditions for neural network approximation
topic Optimization and Control
url https://arxiv.org/abs/2506.17305