Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/24054

Title: Smooth copula-based estimation of the conditional density function with a single covariate
Authors: Janssen, Paul
Swanepoel, Jan
Veraverbeke, Noël
Issue Date: 2017
Abstract: Some recent papers deal with smooth nonparametric estimators for copula functions and copula derivatives. These papers contain results on copula-based Bernstein estimators for conditional distribution functions and related functionals such as regression and quantile functions. The focus in the present paper is on new copula-based smooth Bernstein estimators for the conditional density. Our approach avoids going through separate density estimation of numerator and denominator. Our estimator is defined as a smoother of the copula-based Bernstein estimator of the conditional distribution function. We establish asymptotic properties of bias and variance and discuss the asymptotic mean squared error in terms of the smoothing parameters. We also obtain the asymptotic normality of the new estimator. In a simulation study we show the good performance of the new estimator in comparison with other estimators proposed in the literature.
Notes: Veraverbeke, N (reprint author), Hasselt Univ, Ctr Stat, Agoralaan,Gebouw D, B-3590 Diepenbeek, Belgium. paul.janssen@uhasselt.be; jan.swanepoel@nwu.ac.za; noel.veraverbeke@uhasselt.be
URI: http://hdl.handle.net/1942/24054
DOI: 10.1016/j.jmva.2017.04.008
ISI #: 000405976900003
ISSN: 0047-259X
Category: A1
Type: Journal Contribution
Validation: ecoom, 2018
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
Published version447.7 kBAdobe PDF
Peer-reviewed author version692.65 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.