eprintid: 1748
rev_number: 6
eprint_status: archive
userid: 46
dir: disk0/00/00/17/48
datestamp: 2013-09-17 07:43:09
lastmod: 2015-02-18 11:35:48
status_changed: 2013-09-17 07:43:09
type: article
succeeds: 1737
metadata_visibility: show
creators_name: Gnecco, Giorgio
creators_id: giorgio.gnecco@imtlucca.it
title: Approximation and Estimation Bounds for Subsets of Reproducing Kernel Kreǐn Spaces
ispublished: pub
subjects: QA75
divisions: CSA
full_text_status: none
keywords: Reproducing Kernel Kreǐn Spaces; Estimation error; Approximation error; Rademacher complexity
abstract: Reproducing kernel Kreın spaces are used in learning from data via kernel methods when the kernel is indefinite. In this paper, a characterization of a subset of the unit ball in
such spaces is provided. Conditions are given, under which upper bounds on the estimation error and the approximation error can be applied simultaneously to such a subset. Finally, it is shown that the hyperbolic-tangent kernel and other indefinite kernels satisfy such conditions.
date: 2014-04
date_type: published
publication: Neural Processing Letters
volume: 39
number: 2
publisher: Springer 
pagerange: 137-153
id_number: 10.1007/s11063-013-9294-9
refereed: TRUE
issn: 1370-4621
official_url: http://dx.doi.org/10.1007/s11063-013-9294-9
citation:   Gnecco, Giorgio  Approximation and Estimation Bounds for Subsets of Reproducing Kernel Kreǐn Spaces.  Neural Processing Letters, 39 (2).  pp. 137-153.  ISSN 1370-4621      (2014)