eprintid: 1695
rev_number: 7
eprint_status: archive
userid: 46
dir: disk0/00/00/16/95
datestamp: 2013-09-12 10:56:02
lastmod: 2013-09-16 12:03:00
status_changed: 2013-09-12 10:56:02
type: book_section
metadata_visibility: no_search
creators_name: Gnecco, Giorgio
creators_name: Sanguineti, Marcello
creators_id: giorgio.gnecco@imtlucca.it
creators_id: 
title: Regularization and Suboptimal Solutions in Learning from Data
ispublished: pub
subjects: QA75
divisions: CSA
full_text_status: none
keywords: regularization techniques; accuracy of suboptimal solutions; ill-posedness; inverse problems; weight decay
abstract: Supervised learning from data is investigated from an optimization viewpoint. Ill-posedness issues of the learning problem are discussed and its Tikhonov, Ivanov, Phillips, and Miller regularizations are analyzed. Theoretical features of the optimization problems associated with these regularization techniques and their use in learning tasks are considered. Weight-decay learning is investigated, too. Exploiting properties of the functionals to be minimized in the various regularized problems, estimates are derived on the accuracy of suboptimal solutions formed by linear combinations of n-tuples of computational units, for values of n smaller than the number of data.
date: 2009
date_type: published
series: Studies in Computational Intelligence
number: 247
publisher: Springer 
pagerange: 113-154
id_number: 10.1007/978-3-642-04003-0_6
refereed: TRUE
isbn: 978-3-642-04002-3
book_title: Innovations in Neural Information Paradigms and Applications
official_url: http://dx.doi.org/10.1007/978-3-642-04003-0_6
citation:   Gnecco, Giorgio and Sanguineti, Marcello  Regularization and Suboptimal Solutions in Learning from Data.   In:  Innovations in Neural Information Paradigms and Applications.   Studies in Computational Intelligence  (247).  Springer , pp. 113-154.  ISBN 978-3-642-04002-3      (2009)