Weighted Sobolev Approximation Rates for Neural Networks on Unbounded Domains

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Abdeljawad, Ahmed, Dittrich, Thomas
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912107607359488
author Abdeljawad, Ahmed
Dittrich, Thomas
author_facet Abdeljawad, Ahmed
Dittrich, Thomas
contents In this work, we consider the approximation capabilities of shallow neural networks in weighted Sobolev spaces for functions in the spectral Barron space. The existing literature already covers several cases, in which the spectral Barron space can be approximated well, i.e., without curse of dimensionality, by shallow networks and several different classes of activation function. The limitations of the existing results are mostly on the error measures that were considered, in which the results are restricted to Sobolev spaces over a bounded domain. We will here treat two cases that extend upon the existing results. Namely, we treat the case with bounded domain and Muckenhoupt weights and the case, where the domain is allowed to be unbounded and the weights are required to decay. We first present embedding results for the more general weighted Fourier-Lebesgue spaces in the weighted Sobolev spaces and then we establish asymptotic approximation rates for shallow neural networks that come without curse of dimensionality.
format Preprint
id arxiv_https___arxiv_org_abs_2411_04108
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Weighted Sobolev Approximation Rates for Neural Networks on Unbounded Domains
Abdeljawad, Ahmed
Dittrich, Thomas
Machine Learning
Functional Analysis
41A25, 41A46, 41A30, 46E35, 62M45, 68T05
In this work, we consider the approximation capabilities of shallow neural networks in weighted Sobolev spaces for functions in the spectral Barron space. The existing literature already covers several cases, in which the spectral Barron space can be approximated well, i.e., without curse of dimensionality, by shallow networks and several different classes of activation function. The limitations of the existing results are mostly on the error measures that were considered, in which the results are restricted to Sobolev spaces over a bounded domain. We will here treat two cases that extend upon the existing results. Namely, we treat the case with bounded domain and Muckenhoupt weights and the case, where the domain is allowed to be unbounded and the weights are required to decay. We first present embedding results for the more general weighted Fourier-Lebesgue spaces in the weighted Sobolev spaces and then we establish asymptotic approximation rates for shallow neural networks that come without curse of dimensionality.
title Weighted Sobolev Approximation Rates for Neural Networks on Unbounded Domains
topic Machine Learning
Functional Analysis
41A25, 41A46, 41A30, 46E35, 62M45, 68T05
url https://arxiv.org/abs/2411.04108