Universal approximation property of ODENet and ResNet with a single activation function

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kimura, Masato, Matsui, Kazunori, Mizuno, Yosuke
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914983466500096
author Kimura, Masato
Matsui, Kazunori
Mizuno, Yosuke
author_facet Kimura, Masato
Matsui, Kazunori
Mizuno, Yosuke
contents We study a universal approximation property of ODENet and ResNet. The ODENet is a map from an initial value to the final value of an ODE system in a finite interval. It is considered a mathematical model of a ResNet-type deep learning system. We consider dynamical systems with vector fields given by a single composition of the activation function and an affine mapping, which is the most common choice of the ODENet or ResNet vector field in actual machine learning systems. We show that such an ODENet and ResNet with a restricted vector field can uniformly approximate ODENet with a general vector field.
format Preprint
id arxiv_https___arxiv_org_abs_2410_16709
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Universal approximation property of ODENet and ResNet with a single activation function
Kimura, Masato
Matsui, Kazunori
Mizuno, Yosuke
Machine Learning
Artificial Intelligence
Numerical Analysis
37M15, 41A46, 41A63, 65L12, 68T07, 65P99
We study a universal approximation property of ODENet and ResNet. The ODENet is a map from an initial value to the final value of an ODE system in a finite interval. It is considered a mathematical model of a ResNet-type deep learning system. We consider dynamical systems with vector fields given by a single composition of the activation function and an affine mapping, which is the most common choice of the ODENet or ResNet vector field in actual machine learning systems. We show that such an ODENet and ResNet with a restricted vector field can uniformly approximate ODENet with a general vector field.
title Universal approximation property of ODENet and ResNet with a single activation function
topic Machine Learning
Artificial Intelligence
Numerical Analysis
37M15, 41A46, 41A63, 65L12, 68T07, 65P99
url https://arxiv.org/abs/2410.16709