[Statlist] S'eminaire de Statistique NE

[Werner Stahel]

forwarded from     Messagerie.ISTAT using unine.ch

S'eminaire de Statistique
Institut de Statistique, Universit=E9 de Neuch=E2tel
Salle B29, B=E2timent principal de l'Universit=E9, Neuch=E2tel,
http://www2.unine.ch/statistics
Mardi 7 octobre 2008, 11h00
***************************
Anouar el Ghouch: Department of Econometrics, University of Geneva, Suisse

Local Polynomial Quantile Regression with Parametric Feature

We propose a new approach to conditional quantile function estimation that =
combines both parametric and nonparametric techniques. At each design point=
, a global, possibly incorrect, pilot parametric model is locally adjusted =
through a kernel smoothing fit. The resulting quantile regression estimator=
 behaves like a parametric one when the latter is correct and converges to =
the nonparametric solution as the parametric start deviates from the true u=
nderlying model. We give a Bahadur-type representation of the proposed esti=
mator from which consistency and asymptotic normality are derived under str=
ong mixing assumption. We also discuss numerical implementation and investi=
gate the performance of the estimator via simulations. Finally, we propose =
and numerically study a practical bandwidth selector based on the plug-in p=
rinciple, and we illustrate the methodology on a real data example.

Vous pouvez consulter le programme des s=E9minaires de l'Institut de statis=
tique sur la page http://www2.unine.ch/statistics/page9123.html