Finite sample penalization in adaptive density deconvolution

dc.creatorComte, Fabienne
dc.creatorRozenholc, Yves
dc.creatorTaupin, Marie-Luce
dc.date2006-01-05
dc.date.accessioned2026-07-07T09:19:40Z
dc.date.available2026-07-07T09:19:40Z
dc.descriptionWe consider the problem of estimating the density $g$ of identically distributed variables $X\_i$, from a sample $Z\_1, ..., Z\_n$ where $Z\_i=X\_i+σε\_i$, $i=1, ..., n$ and $σε\_i$ is a noise independent of $X\_i$ with known density $ σ^{-1}f\_ε(./σ)$. We generalize adaptive estimators, constructed by a model selection procedure, described in Comte et al. (2005). We study numerically their properties in various contexts and we test their robustness. Comparisons are made with respect to deconvolution kernel estimators, misspecification of errors, dependency,... It appears that our estimation algorithm, based on a fast procedure, performs very well in all contexts.
dc.identifierhttps://arxiv.org/abs/math/0601098
dc.identifierhttp://arxiv.org/abs/math/0601098
dc.identifierJournal of Statistical Computation and Simulation 77, 11 (2007) 977-1000
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/154476
dc.subjectStatistics Theory
dc.subjectPrimary 62G07. Secondary 62G20
dc.titleFinite sample penalization in adaptive density deconvolution
dc.typetext

Files

Collections