Exponential convergence of the deep neural network approximation for analytic functions

DOI10.1007/S11425-018-9387-XzbMATH Open1475.65007arXiv1807.00297OpenAlexW2963172624WikidataQ115602254 ScholiaQ115602254MaRDI QIDQ1989902FDOQ1989902

Authors: Qingcan Wang, Weinan E 

Publication date: 29 October 2018

Published in: Science China. Mathematics (Search for Journal in Brave)

Abstract: We prove that for analytic functions in low dimension, the convergence rate of the deep neural network approximation is exponential.

Full work available at URL: https://arxiv.org/abs/1807.00297

zbMATH Keywords

approximation theoryneural networksanalytic functions

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Algorithms for approximation of functions (65D15)

Cites Work

• Universal approximation bounds for superpositions of a sigmoidal function
• Approximation by superpositions of a sigmoidal function
• Error bounds for approximations with deep ReLU networks

Cited In (42)

• Constructive deep ReLU neural network approximation
• Deep Neural Networks for Solving Large Linear Systems Arising from High-Dimensional Problems
• Exponential ReLU DNN expression of holomorphic maps in high dimension
• N-adaptive Ritz method: a neural network enriched partition of unity for boundary value problems
• PDE-Net 2.0: learning PDEs from data with a numeric-symbolic hybrid deep network
• Rectified deep neural networks overcome the curse of dimensionality for nonsmooth value functions in zero-sum games of nonlinear stiff systems
• A note on the applications of one primary function in deep neural networks
• MgNet: a unified framework of multigrid and convolutional neural network
• Optimization of random feature method in the high-precision regime
• Collocation approximation by deep neural ReLU networks for parametric and stochastic PDEs with lognormal inputs
• Computation and learning in high dimensions. Abstracts from the workshop held August 1--7, 2021 (hybrid meeting)
• Deep Network Approximation for Smooth Functions
• Structure probing neural network deflation
• Selection dynamics for deep neural networks
• On the approximation of functions by tanh neural networks
• Deep ReLU neural networks in high-dimensional approximation
• Deep network with approximation error being reciprocal of width to power of square root of depth
• Analytic continuation of noisy data using Adams Bashforth residual neural network
• SelectNet: self-paced learning for high-dimensional partial differential equations
• Deep ReLU Networks Overcome the Curse of Dimensionality for Generalized Bandlimited Functions
• Improved training of physics-informed neural networks for parabolic differential equations with sharply perturbed initial conditions
• Deep ReLU neural network approximation in Bochner spaces and applications to parametric PDEs
• Convergence of deep convolutional neural networks
• Overall error analysis for the training of deep neural networks via stochastic gradient descent with random initialisation
• Nonlinear approximation and (deep) ReLU networks
• ReLU deep neural networks from the hierarchical basis perspective
• Nonlinear approximation via compositions
• On mathematical modeling in image reconstruction and beyond
• Better approximations of high dimensional smooth functions by deep neural networks with rectified power units
• Neural network approximation
• Sparse approximation of triangular transports. I: The finite-dimensional case
• Computing ground states of Bose-Einstein condensation by normalized deep neural network
• The gap between theory and practice in function approximation with deep neural networks
• Optimal approximation rate of ReLU networks in terms of width and depth
• Solving PDEs on unknown manifolds with machine learning
• Simultaneous neural network approximation for smooth functions
• Non-homogeneous Poisson process intensity modeling and estimation using measure transport
• Deep Neural Networks with ReLU-Sine-Exponential Activations Break Curse of Dimensionality in Approximation on Hölder Class
• Deep ReLU networks and high-order finite element methods
• Neural network expression rates and applications of the deep parametric PDE method in counterparty credit risk
• Response to stress via underlying deep gene regulation networks
• Full error analysis for the training of deep neural networks

This page was built for publication: Exponential convergence of the deep neural network approximation for analytic functions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1989902)