Confidence bands for regression curves and their first p derivatives are obtained via local pth order polynomial estimation. The method allows for multiparameter local likelihood estimation as well as other unbiased estimating equations. As an alternative to the confidence bands obtained by asymptotic distribution theory, we also study smoothed bootstrap confidence bands. Simulations illustrate the finite sample properties of the methodology.
Publié le : 2003-12-14
Classification:
Confidence band,
lack-of-fit test,
local estimation equations,
local polynomial esimation,
multiparameter local likelihood,
one-step bootstrap,
smoothed bootstrap,
62G15,
62E20,
62G09
@article{1074290329,
author = {Claeskens, Gerda and Keilegom, Ingrid van},
title = {Bootstrap confidence bands for regression curves and their derivatives},
journal = {Ann. Statist.},
volume = {31},
number = {1},
year = {2003},
pages = { 1852-1884},
language = {en},
url = {http://dml.mathdoc.fr/item/1074290329}
}
Claeskens, Gerda; Keilegom, Ingrid van. Bootstrap confidence bands for regression curves and their derivatives. Ann. Statist., Tome 31 (2003) no. 1, pp. 1852-1884. http://gdmltest.u-ga.fr/item/1074290329/