In this paper, an entropy approach is proposed to establish rates of convergence for estimators of a regression function. General regression problems are considered, with linear regression, splines and isotonic regression as special cases. The estimation methods studied are least squares, least absolute deviations and penalized least squares. Common features of these methods and various regression problems are highlighted.
Publié le : 1990-06-14
Classification:
Empirical processes,
entropy,
least absolute deviations,
(penalized) least squares,
rates of convergence,
60B10,
60G50,
62J99
@article{1176347632,
author = {van de Geer, Sara},
title = {Estimating a Regression Function},
journal = {Ann. Statist.},
volume = {18},
number = {1},
year = {1990},
pages = { 907-924},
language = {en},
url = {http://dml.mathdoc.fr/item/1176347632}
}
van de Geer, Sara. Estimating a Regression Function. Ann. Statist., Tome 18 (1990) no. 1, pp. 907-924. http://gdmltest.u-ga.fr/item/1176347632/