NONPARAMETRIC-ESTIMATION VIA EMPIRICAL RISK MINIMIZATION

Authors
LUGOSI G ZEGER K
Citation
G. Lugosi et K. Zeger, NONPARAMETRIC-ESTIMATION VIA EMPIRICAL RISK MINIMIZATION, IEEE transactions on information theory, 41(3), 1995, pp. 677-687
Citations number
78
Categorie Soggetti
Information Science & Library Science","Engineering, Eletrical & Electronic
Journal title
IEEE transactions on information theory → ACNP
ISSN journal
00189448
Volume
41
Issue
3
Year of publication
1995
Pages
677 - 687
Database
ISI
SICI code
0018-9448(1995)41:3<677:NVERM>2.0.ZU;2-U
Abstract
A general notion of universal consistency of nonparametric estimators is introduced that applies to regression estimation, conditional media n estimation, curve fitting, pattern recognition, and learning concept s. General methods for proving consistency of estimators based on mini mizing the empirical error are shown, In particular, distribution-free almost sure consistency of neural network estimates and generalized l inear estimators is established.