Error bounds for approximation with neural networks

DOI10.1006/jath.2001.3613zbMath1004.41007MaRDI QIDQ5959036

Martin Burger, Andreas Neubauer

Publication date: 28 April 2002

Published in: Journal of Approximation Theory (Search for Journal in Brave)

Full work available at URL: https://semanticscholar.org/paper/48d823e894306e4b0cff3d54eee343da415474e7

zbMATH Keywords

neural networks; error bounds; approximation order

Mathematics Subject Classification ID

41A25: Rate of convergence, degree of approximation

65D15: Algorithms for approximation of functions

41A46: Approximation by arbitrary nonlinear expressions; widths and entropy

33F05: Numerical approximation and evaluation of special functions

Related Items

Consistency of Ridge Function Fields for Varying Nonparametric Regression, Estimation of approximating rate for neural network in \(L^p_w\) spaces, Suboptimal solutions to dynamic optimization problems via approximations of the policy functions, The ridge function representation of polynomials and an application to neural networks, Nonlinear function approximation: computing smooth solutions with an adaptive greedy algorithm, Complexity of Gaussian-radial-basis networks approximating smooth functions, Learning a function from noisy samples at a finite sparse set of points, Approximation by neural networks with a bounded number of nodes at each level, Approximation by ridge function fields over compact sets, Almost optimal estimates for approximation and learning by radial basis function networks

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Generalization bounds for function approximation from scattered noisy data
• Multilayer feedforward networks are universal approximators
• Approximation properties of a multilayered feedforward artificial neural network
• Degree of approximation by neural and translation networks with a single hidden layer
• Approximation by radial basis functions with finitely many centers
• Convergence rates of certain approximate solutions to Fredholm integral equations of the first kind
• Approximation by Ridge Functions and Neural Networks
• Universal approximation bounds for superpositions of a sigmoidal function
• Training neural networks with noisy data as an ill-posed problem