Relative error prediction in nonparametric deconvolution regression model

Abstract : In this paper, we studied an alternative estimator of the regression function when the covariates are observed with error. It is based on the minimization of the relative mean squared error. We obtain expressions for its asymptotic bias and variance together with an asymptotic normality result. Our technique is illustrated on simulation studies. Numerical results suggest that the studied estimator can lead to tangible improvements in prediction over the usual kernel deconvolution regression estimator, particularly in the presence of several outliers in the dataset.
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01824274
Contributor : Isabelle Celet <>
Submitted on : Wednesday, June 27, 2018 - 9:28:04 AM
Last modification on : Thursday, April 11, 2019 - 9:25:01 AM

Links full text

Identifiers

Collections

Citation

Baba Thiam. Relative error prediction in nonparametric deconvolution regression model. Statistica Neerlandica, Wiley, 2018, ⟨10.1111/stan.12135⟩. ⟨hal-01824274⟩

Share

Metrics

Record views

53