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. 113154. ISBN 9783642040023 (2009)
This is the latest version of this item.
Abstract
Supervised learning from data is investigated from an optimization viewpoint. Illposedness 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. Weightdecay 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 ntuples of computational units, for values of n smaller than the number of data.
Item Type:  Book Section 

Identification Number:  10.1007/9783642040030_6 
Uncontrolled Keywords:  regularization techniques; accuracy of suboptimal solutions; illposedness; inverse problems; weight decay 
Subjects:  Q Science > QA Mathematics > QA75 Electronic computers. Computer science 
Research Area:  Computer Science and Applications 
Depositing User:  Giorgio Gnecco 
Date Deposited:  17 Sep 2013 13:10 
Last Modified:  17 Sep 2013 13:10 
URI:  http://eprints.imtlucca.it/id/eprint/1776 
Available Versions of this Item

Regularization and Suboptimal Solutions in Learning from Data. (deposited 12 Sep 2013 10:56)
 Regularization and Suboptimal Solutions in Learning from Data. (deposited 17 Sep 2013 13:10) [Currently Displayed]
Actions (login required)
Edit Item 