Regularized greedy algorithms for network training with data noise

DOI10.1007/S00607-004-0081-3MaRDI QIDQ1774153zbMATH OpenOpenAlexFDO

Authors Martin Burger, Andreas Hofinger

Publication date 29 April 2005

Published in Computing (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1007/s00607-004-0081-3

zbMATH Keywords

data noiseGreedy approximationneural networksregularizationtraining

Mathematics Subject Classification ID

Rate of convergence, degree of approximation (41A25) Best approximation, Chebyshev systems (41A50) Neural networks for/in biological studies, artificial life and related topics (92B20)

Review text

The aim of this paper is to construct a modified greedy algorithm applicable to an ill-posed function approximation problem in presence of data noise. The authors provide a detailed convergence analysis of the algorithm and discuss the choice of the iteration parameters. Applications to neural networks are presented.

Recommendations

• Training neural networks with noisy data as an ill-posed problem
• Rough Sets and Current Trends in Computing
• Nonlinear function approximation: computing smooth solutions with an adaptive greedy algorithm
• Approximation and learning by greedy algorithms
• A simple lemma on greedy approximation in Hilbert space and convergence rates for projection pursuit regression and neural network training

Cites work

• A note on error bounds for approximation in inner product spaces
• A simple lemma on greedy approximation in Hilbert space and convergence rates for projection pursuit regression and neural network training
• scientific article; zbMATH DE number 3834323 (Why is no real title available?)
• scientific article; zbMATH DE number 812882 (Why is no real title available?)
• scientific article; zbMATH DE number 936298 (Why is no real title available?)
• scientific article; zbMATH DE number 3360568 (Why is no real title available?)
• Multilayer feedforward networks are universal approximators
• Nonlinear black-box modeling in system identification: A unified overview
• Regularized data-driven construction of fuzzy controllers
• Regularized greedy algorithms for network training with data noise
• Training neural networks with noisy data as an ill-posed problem
• Universal approximation bounds for superpositions of a sigmoidal function

Cited in

(10)

• Learning a function from noisy samples at a finite sparse set of points
• Regularized greedy algorithms for network training with data noise
• Local greedy approximation for nonlinear regression and neural network training.
• Neural network training using \(\ell_1\)-regularization and bi-fidelity data
• A note on greedy algorithm
• scientific article; zbMATH DE number 1304397 (Why is no real title available?)
• Rough Sets and Current Trends in Computing
• Node perturbation learning without noiseless baseline
• Training neural networks with noisy data as an ill-posed problem
• Nonlinear function approximation: computing smooth solutions with an adaptive greedy algorithm

This page was built for publication: Regularized greedy algorithms for network training with data noise

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1774153)