[Résultats d'admissibilité dans une classe de fonctions de perte équilibrées dans un modèle de régression fonctionnelle]
On considère le problème de l'estimation non paramétrique dans un modèle de régression fonctionnelle , où Y est une variable aléatoire réelle et X est une variable fonctionnelle à valeurs dans un espace semi-métrique. Le but de cette note est de trouver les conditions d'admissibilité des estimateurs de type Stein de ce modèle dans une classe de fonctions de perte équilibrées. Notre méthode consiste à comparer le risque avec celui obtenu dans le cas d'une perte quadratique.
We consider the problem of the nonparametric estimation in a functional regression model , with Y a real random variable response and X representing a functional variable taking values in a semi-metric space. The aim of this note is to find conditions of admissibility of Stein-type estimators of such a model under a class of balanced loss functions. Our method is to compare the risk with that obtained in the case of a quadratic loss.
Accepté le :
Publié le :
@article{CRMATH_2019__357_11-12_912_0,
author = {Djerfi, Kouider and Madani, Fethi and Ouassou, Idir},
title = {Admissibility results under some balanced loss functions for a functional regression model},
journal = {Comptes Rendus. Math\'ematique},
pages = {912--916},
publisher = {Elsevier},
volume = {357},
number = {11-12},
year = {2019},
doi = {10.1016/j.crma.2019.10.012},
language = {en},
url = {http://www.numdam.org/articles/10.1016/j.crma.2019.10.012/}
}
TY - JOUR AU - Djerfi, Kouider AU - Madani, Fethi AU - Ouassou, Idir TI - Admissibility results under some balanced loss functions for a functional regression model JO - Comptes Rendus. Mathématique PY - 2019 SP - 912 EP - 916 VL - 357 IS - 11-12 PB - Elsevier UR - http://www.numdam.org/articles/10.1016/j.crma.2019.10.012/ DO - 10.1016/j.crma.2019.10.012 LA - en ID - CRMATH_2019__357_11-12_912_0 ER -
%0 Journal Article %A Djerfi, Kouider %A Madani, Fethi %A Ouassou, Idir %T Admissibility results under some balanced loss functions for a functional regression model %J Comptes Rendus. Mathématique %D 2019 %P 912-916 %V 357 %N 11-12 %I Elsevier %U http://www.numdam.org/articles/10.1016/j.crma.2019.10.012/ %R 10.1016/j.crma.2019.10.012 %G en %F CRMATH_2019__357_11-12_912_0
Djerfi, Kouider; Madani, Fethi; Ouassou, Idir. Admissibility results under some balanced loss functions for a functional regression model. Comptes Rendus. Mathématique, Tome 357 (2019) no. 11-12, pp. 912-916. doi : 10.1016/j.crma.2019.10.012. http://www.numdam.org/articles/10.1016/j.crma.2019.10.012/
[1] Local smoothing regression with functional data, Comput. Stat., Volume 22 (2007) no. 22, pp. 353-369
[2] Consistency of the regression estimator with functional data under long memory conditions, Stat. Probab. Lett., Volume 78 (2008) no. 8, pp. 1043-1049 | DOI
[3] Kernel-based functional principal components, Stat. Probab. Lett., Volume 48 (2000) no. 4, pp. 335-345
[4] Linear Processes in Function Space: Theory and Application, Lecture Notes in Statistics, vol. 149, Springer-Verlag, New York, 2002
[5] Generalizations of James–Stein estimators under spherical symmetry, Ann. Stat., Volume 19 (1991) no. 3, pp. 1639-1650
[6] Shrinkage estimators of the location parameter for certain spherically symmetric distributions, Ann. Inst. Stat. Math., Volume 45 (1991) no. 3, pp. 551-563
[7] Shrinkage estimators under spherical symmetry for the general linear model, J. Multivar. Anal., Volume 52 (1995) no. 2, pp. 338-351
[8] Robust shrinkage estimators of the location parameter for elliptically symmetric distributions, J. Multivar. Anal., Volume 29 (1989), pp. 39-52
[9] A partial overview of the theory of statistics with functional data, J. Stat. Plan. Inference, Volume 147 (2014), pp. 1-23
[10] On estimation with balanced loss functions, Stat. Probab. Lett., Volume 45 (1999), pp. 97-101
[11] Nonparametric Functional Data Analysis. Theory and Practice, Springer-Verlag, New York, 2006
[12] Estimation of a loss function for spherically symmetric distributions in the general linear model, Ann. Stat., Volume 23 (1995) no. 2, pp. 571-592
[13] On improved loss estimation for shrinkage estimators, Stat. Sci., Volume 27 (2012) no. 1, pp. 61-81
[14] An introduction to recent advances in high/infinite dimensional statistics, J. Multivar. Anal., Volume 146 (2016), pp. 1-6 | DOI
[15] Estimation with quadratic loss, Statistical Laboratory of the University of California, Berkeley, 20 June–30 July, 1960, University of California Press, Berkeley, CA, USA (1961), pp. 361-379
[16] On estimation with weighted balanced-type loss function, Stat. Probab. Lett., Volume 76 (2006), pp. 773-780
[17] Nonparametric regression for functional data: automatic smoothing parameter selection, J. Stat. Plan. Inference, Volume 137 (2007) no. 9, pp. 2784-2801
[18] Some tools for functional data analysis, J. R. Stat. Soc. B, Volume 53 (1991) no. 3, pp. 539-572 (with discussion)
[19] Functional Data Analysis, Springer Series in Statistics, Springer-Verlag, New York, 1997
[20] Applied Functional Data Analysis: Methods and Case Studies, Springer Series in Statistics, Springer-Verlag, New York, 2002
[21] Inadmissibility of the usual estimator for the mean of a multivariate normal distribution, Statistical Laboratory of the University of California, Berkeley and Los Angeles, December, 1954 and July–August, 1955, University of California Press (1956), pp. 197-206
[22] Bayesian and non-Bayesian estimation using balanced loss functions, Statistical Decision Theory and Related Topics V, Springer, New York, 1994, pp. 337-390
Cité par Sources :
