On the Dimension-Free Approximation of Deep Neural Networks for Symmetric Korobov Functions

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lu, Yulong, Mao, Tong, Xu, Jinchao, Yang, Yahong
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908656656711680
author Lu, Yulong
Mao, Tong
Xu, Jinchao
Yang, Yahong
author_facet Lu, Yulong
Mao, Tong
Xu, Jinchao
Yang, Yahong
contents Deep neural networks have been widely used as universal approximators for functions with inherent physical structures, including permutation symmetry. In this paper, we construct symmetric deep neural networks to approximate symmetric Korobov functions and prove that both the convergence rate and the constant prefactor scale at most polynomially with respect to the ambient dimension. This represents a substantial improvement over prior approximation guarantees that suffer from the curse of dimensionality. Building on these approximation bounds, we further derive a generalization-error rate for learning symmetric Korobov functions whose leading factors likewise avoid the curse of dimensionality.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12398
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On the Dimension-Free Approximation of Deep Neural Networks for Symmetric Korobov Functions
Lu, Yulong
Mao, Tong
Xu, Jinchao
Yang, Yahong
Machine Learning
Numerical Analysis
Deep neural networks have been widely used as universal approximators for functions with inherent physical structures, including permutation symmetry. In this paper, we construct symmetric deep neural networks to approximate symmetric Korobov functions and prove that both the convergence rate and the constant prefactor scale at most polynomially with respect to the ambient dimension. This represents a substantial improvement over prior approximation guarantees that suffer from the curse of dimensionality. Building on these approximation bounds, we further derive a generalization-error rate for learning symmetric Korobov functions whose leading factors likewise avoid the curse of dimensionality.
title On the Dimension-Free Approximation of Deep Neural Networks for Symmetric Korobov Functions
topic Machine Learning
Numerical Analysis
url https://arxiv.org/abs/2511.12398