Saved in:
Bibliographic Details
Main Authors: García, Jonathan, Petersen, Philipp
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2412.07312
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912182617243648
author García, Jonathan
Petersen, Philipp
author_facet García, Jonathan
Petersen, Philipp
contents We prove that a classifier with a Barron-regular decision boundary can be approximated with a rate of high polynomial degree by ReLU neural networks with three hidden layers when a margin condition is assumed. In particular, for strong margin conditions, high-dimensional discontinuous classifiers can be approximated with a rate that is typically only achievable when approximating a low-dimensional smooth function. We demonstrate how these expression rate bounds imply fast-rate learning bounds that are close to $n^{-1}$ where $n$ is the number of samples. In addition, we carry out comprehensive numerical experimentation on binary classification problems with various margins. We study three different dimensions, with the highest dimensional problem corresponding to images from the MNIST data set.
format Preprint
id arxiv_https___arxiv_org_abs_2412_07312
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle High-dimensional classification problems with Barron regular boundaries under margin conditions
García, Jonathan
Petersen, Philipp
Machine Learning
Probability
68T05, 62C20, 41A25, 41A46
We prove that a classifier with a Barron-regular decision boundary can be approximated with a rate of high polynomial degree by ReLU neural networks with three hidden layers when a margin condition is assumed. In particular, for strong margin conditions, high-dimensional discontinuous classifiers can be approximated with a rate that is typically only achievable when approximating a low-dimensional smooth function. We demonstrate how these expression rate bounds imply fast-rate learning bounds that are close to $n^{-1}$ where $n$ is the number of samples. In addition, we carry out comprehensive numerical experimentation on binary classification problems with various margins. We study three different dimensions, with the highest dimensional problem corresponding to images from the MNIST data set.
title High-dimensional classification problems with Barron regular boundaries under margin conditions
topic Machine Learning
Probability
68T05, 62C20, 41A25, 41A46
url https://arxiv.org/abs/2412.07312