Neural Networks for Singular Perturbations

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Opschoor, Joost A. A., Schwab, Christoph, Xenophontos, Christos
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866929207674667008
author Opschoor, Joost A. A.
Schwab, Christoph
Xenophontos, Christos
author_facet Opschoor, Joost A. A.
Schwab, Christoph
Xenophontos, Christos
contents We prove deep neural network (DNN for short) expressivity rate bounds for solution sets of a model class of singularly perturbed, elliptic two-point boundary value problems, in Sobolev norms, on the bounded interval $(-1,1)$. We assume that the given source term and reaction coefficient are analytic in $[-1,1]$. We establish expression rate bounds in Sobolev norms in terms of the NN size which are uniform with respect to the singular perturbation parameter for several classes of DNN architectures. In particular, ReLU NNs, spiking NNs, and $\tanh$- and sigmoid-activated NNs. The latter activations can represent ``exponential boundary layer solution features'' explicitly, in the last hidden layer of the DNN, i.e. in a shallow subnetwork, and afford improved robust expression rate bounds in terms of the NN size. We prove that all DNN architectures allow robust exponential solution expression in so-called `energy' as well as in `balanced' Sobolev norms, for analytic input data.
format Preprint
id arxiv_https___arxiv_org_abs_2401_06656
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Neural Networks for Singular Perturbations
Opschoor, Joost A. A.
Schwab, Christoph
Xenophontos, Christos
Numerical Analysis
Machine Learning
34B08, 34D15, 65L11
We prove deep neural network (DNN for short) expressivity rate bounds for solution sets of a model class of singularly perturbed, elliptic two-point boundary value problems, in Sobolev norms, on the bounded interval $(-1,1)$. We assume that the given source term and reaction coefficient are analytic in $[-1,1]$. We establish expression rate bounds in Sobolev norms in terms of the NN size which are uniform with respect to the singular perturbation parameter for several classes of DNN architectures. In particular, ReLU NNs, spiking NNs, and $\tanh$- and sigmoid-activated NNs. The latter activations can represent ``exponential boundary layer solution features'' explicitly, in the last hidden layer of the DNN, i.e. in a shallow subnetwork, and afford improved robust expression rate bounds in terms of the NN size. We prove that all DNN architectures allow robust exponential solution expression in so-called `energy' as well as in `balanced' Sobolev norms, for analytic input data.
title Neural Networks for Singular Perturbations
topic Numerical Analysis
Machine Learning
34B08, 34D15, 65L11
url https://arxiv.org/abs/2401.06656