Theory-to-Practice Gap for Neural Networks and Neural Operators

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Grohs, Philipp, Lanthaler, Samuel, Trautner, Margaret
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908280341659648
author Grohs, Philipp
Lanthaler, Samuel
Trautner, Margaret
author_facet Grohs, Philipp
Lanthaler, Samuel
Trautner, Margaret
contents This work studies the sampling complexity of learning with ReLU neural networks and neural operators. For mappings belonging to relevant approximation spaces, we derive upper bounds on the best-possible convergence rate of any learning algorithm, with respect to the number of samples. In the finite-dimensional case, these bounds imply a gap between the parametric and sampling complexities of learning, known as the \emph{theory-to-practice gap}. In this work, a unified treatment of the theory-to-practice gap is achieved in a general $L^p$-setting, while at the same time improving available bounds in the literature. Furthermore, based on these results the theory-to-practice gap is extended to the infinite-dimensional setting of operator learning. Our results apply to Deep Operator Networks and integral kernel-based neural operators, including the Fourier neural operator. We show that the best-possible convergence rate in a Bochner $L^p$-norm is bounded by Monte-Carlo rates of order $1/p$.
format Preprint
id arxiv_https___arxiv_org_abs_2503_18219
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Theory-to-Practice Gap for Neural Networks and Neural Operators
Grohs, Philipp
Lanthaler, Samuel
Trautner, Margaret
Machine Learning
Functional Analysis
This work studies the sampling complexity of learning with ReLU neural networks and neural operators. For mappings belonging to relevant approximation spaces, we derive upper bounds on the best-possible convergence rate of any learning algorithm, with respect to the number of samples. In the finite-dimensional case, these bounds imply a gap between the parametric and sampling complexities of learning, known as the \emph{theory-to-practice gap}. In this work, a unified treatment of the theory-to-practice gap is achieved in a general $L^p$-setting, while at the same time improving available bounds in the literature. Furthermore, based on these results the theory-to-practice gap is extended to the infinite-dimensional setting of operator learning. Our results apply to Deep Operator Networks and integral kernel-based neural operators, including the Fourier neural operator. We show that the best-possible convergence rate in a Bochner $L^p$-norm is bounded by Monte-Carlo rates of order $1/p$.
title Theory-to-Practice Gap for Neural Networks and Neural Operators
topic Machine Learning
Functional Analysis
url https://arxiv.org/abs/2503.18219