On bandwidth selection in partial linear regression models under dependence

We obtain the expression of an asymptotically optimal bandwidth for a semiparametric least-squares estimator of [beta] in the model y=xT[beta]+m(t)+[var epsilon], where x is random, t is fixed, m is unknown and [var epsilon] is strong mixing. The selection method is based on second-order approximations for the variance and bias. Asymptotic normality is also established.

MoreLess

Year of publication:	2002
Authors:	Aneiros-Pérez, Germán
Published in:	Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 57.2002, 4, p. 393-401
Publisher:	Elsevier
Keywords:	Partial linear models Kernel smoothing Bandwidth selection Mixing

More details

Type of publication:	Article
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10005319567