ADAPTIVE DECONVOLUTION OF LINEAR FUNCTIONALS ON THE NONNEGATIVE REAL LINE

Abstract : In this paper we consider the convolution model Z = X + Y with X of unknown density f , independent of Y , when both random variables are nonnegative. Our goal is to estimate linear functionals of f such as f for a known function ψ assuming that the distribution of Y is known and only Z is observed. We propose an estimator of f based on a projection estimator of f on Laguerre spaces, present upper bounds on the quadratic risk and derive the rate of convergence in function of the smoothness of f, g and ψ. Then we propose a nonparametric data driven strategy, inspired Goldenshluger and Lepski (2011) method to select a relevant projection space. This methodology is then adapted to pointwise estimation of f. We illustrate the good performance of the new method through simulations. We also test a new approach for choosing the tuning parameter in Goldenshluger-Lepski data driven estimators following ideas developed in Lacour and Massart (2015).
Document type :
Journal articles
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01195711
Contributor : Gwennaëlle Mabon <>
Submitted on : Wednesday, March 30, 2016 - 9:36:48 AM
Last modification on : Thursday, April 11, 2019 - 4:02:09 PM
Document(s) archivé(s) le : Monday, November 14, 2016 - 8:58:03 AM

File

Mabon_linear_functional_v2.pdf
Files produced by the author(s)

Identifiers

Collections

Citation

Gwennaëlle Mabon. ADAPTIVE DECONVOLUTION OF LINEAR FUNCTIONALS ON THE NONNEGATIVE REAL LINE. Journal of Statistical Planning and Inference, Elsevier, 2016, 178, pp.1-23. ⟨10.1016/j.jspi.2016.04.006⟩. ⟨hal-01195711v2⟩

Share

Metrics

Record views

182

Files downloads

95