Fredholm integral equations for function approximation and the training of neural networks

Patrick Gelß, Aizhan Issagali, Ralf Kornhuber

Publication date: 20 December 2024

Published in: SIAM Journal on Mathematics of Data Science (Search for Journal in Brave)

zbMATH Keywords

Tikhonov regularization function approximation Fredholm integral equations of the first kind Ritz-Galerkin methods training of neural networks tensor trains

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Numerical methods for integral equations (65R20) Algorithms for approximation of functions (65D15)

Cites Work

Unnamed Item
Unnamed Item
Unnamed Item
Tensor Decompositions and Applications
Tensor-Train Decomposition
Backpropagation and stochastic gradient descent method
Linear integral equations.
A note on the Poincaré inequality for convex domains
The expression of a tensor or a polyadic as a sum of products.
Nearest-neighbor interaction systems in the tensor-train format
On minimal subspaces in tensor representations
A continuous analogue of the tensor-train decomposition
Non-convergence of stochastic gradient descent in the training of deep neural networks
Mean-field Langevin dynamics and energy landscape of neural networks
Randomized Newton's method for solving differential equations based on the neural network discretization
Stable als approximation in the TT-format for rank-adaptive tensor completion
On manifolds of tensors of fixed TT-rank
On Local Convergence of Alternating Schemes for Optimization of Convex Problems in the Tensor Train Format
The Alternating Linear Scheme for Tensor Optimization in the Tensor Train Format
Alternating Minimal Energy Methods for Linear Systems in Higher Dimensions
Tensor Spaces and Numerical Tensor Calculus
Breaking the Curse of Dimensionality, Or How to Use SVD in Many Dimensions
Trust Region Methods
A mean field view of the landscape of two-layer neural networks
Trainability and Accuracy of Artificial Neural Networks: An Interacting Particle System Approach
Approximation by superpositions of a sigmoidal function
Globally Convergent Multilevel Training of Deep Residual Networks

This page was built for publication: Fredholm integral equations for function approximation and the training of neural networks