Learning improvement of neural networks used in structural optimization

Lagaros, ND; Papadrakakis, M

dc.contributor.author	Lagaros, ND	en
dc.contributor.author	Papadrakakis, M	en
dc.date.accessioned	2014-03-01T01:20:43Z
dc.date.available	2014-03-01T01:20:43Z
dc.date.issued	2004	en
dc.identifier.issn	0965-9978	en
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/16021
dc.subject	Evolution strategies	en
dc.subject	Ill-conditioning	en
dc.subject	Neural networks	en
dc.subject	Structural optimization	en
dc.subject.classification	Computer Science, Interdisciplinary Applications	en
dc.subject.classification	Computer Science, Software Engineering	en
dc.subject.other	Algorithms	en
dc.subject.other	Approximation theory	en
dc.subject.other	Backpropagation	en
dc.subject.other	Computational methods	en
dc.subject.other	Finite element method	en
dc.subject.other	Fracture mechanics	en
dc.subject.other	Learning systems	en
dc.subject.other	Matrix algebra	en
dc.subject.other	Neural networks	en
dc.subject.other	Numerical methods	en
dc.subject.other	Problem solving	en
dc.subject.other	Statistical methods	en
dc.subject.other	Structural optimization	en
dc.subject.other	Vectors	en
dc.subject.other	Evolution strategies	en
dc.subject.other	III-conditioning	en
dc.subject.other	Software engineering	en
dc.title	Learning improvement of neural networks used in structural optimization	en
heal.type	journalArticle	en
heal.identifier.primary	10.1016/S0965-9978(03)00112-1	en
heal.identifier.secondary	http://dx.doi.org/10.1016/S0965-9978(03)00112-1	en
heal.language	English	en
heal.publicationDate	2004	en
heal.abstract	The performance of feed-forward neural networks can be substantially impaired by the ill-conditioning of the corresponding Jacobian matrix. Ill-conditioning appearing in feed-forward learning process is related to the properties of the activation function used. It will be shown that the performance of the network training can be improved using an adaptive activation function with a properly updated gain parameter during the learning process. The efficiency of the proposed adaptive procedure is examined in structural optimization problems where a trained neural network is used to replace the structural analysis phase and capture the necessary data for the optimizer. The optimizer used in this study is an algorithm based on evolution strategies. (C) 2003 Elsevier Ltd. All rights reserved.	en
heal.publisher	ELSEVIER SCI LTD	en
heal.journalName	Advances in Engineering Software	en
dc.identifier.doi	10.1016/S0965-9978(03)00112-1	en
dc.identifier.isi	ISI:000188879100002	en
dc.identifier.volume	35	en
dc.identifier.issue	1	en
dc.identifier.spage	9	en
dc.identifier.epage	25	en