eprintid: 1669
rev_number: 9
eprint_status: archive
userid: 46
dir: disk0/00/00/16/69
datestamp: 2013-09-10 15:08:56
lastmod: 2013-09-16 12:03:00
status_changed: 2013-09-10 15:08:56
type: book_section
metadata_visibility: show
creators_name: Gnecco, Giorgio
creators_name: Kůrková, Věra
creators_name: Sanguineti, Marcello
creators_id: giorgio.gnecco@imtlucca.it
creators_id: 
creators_id: 
title: Some Comparisons of Model Complexity in Linear and Neural-Network Approximation
ispublished: pub
subjects: QA75
divisions: CSA
full_text_status: none
pres_type: paper
note: Proceedings of the 20th International Conference, Thessaloniki, Greece, September 15-18, 2010, Part III

abstract: Capabilities of linear and neural-network models are compared from the point of view of requirements on the growth of model complexity with an increasing accuracy of approximation. Upper bounds on worst-case errors in approximation by neural networks are compared with lower bounds on these errors in linear approximation. The bounds are formulated in terms of singular numbers of certain operators induced by computational units and high-dimensional volumes of the domains of the functions to be approximated.
date: 2010
date_type: published
series: Lecture Notes in Computer Science
number: 6354
publisher: Springer
pagerange: 358-367
refereed: TRUE
isbn: 978-3-642-15825-4
book_title: Artificial Neural Networks – ICANN 2010
official_url: http://dx.doi.org/10.1007/978-3-642-15825-4_48
citation:   Gnecco, Giorgio and Kůrková, Věra and Sanguineti, Marcello  Some Comparisons of Model Complexity in Linear and Neural-Network Approximation.   In:  Artificial Neural Networks – ICANN 2010.   Lecture Notes in Computer Science  (6354).  Springer, pp. 358-367.  ISBN 978-3-642-15825-4      (2010)