Bandwidth selection for kernel density estimation with doubly truncated data

Several bandwidth selection procedures for kernel density estimation of a random variable that is sampled under random double truncation are introduced and compared. The motivation is based on the fact that this type of incomplete data is often encountered in astronomy and medicine. The considered bandwidth selection procedures are appropriate modifications of the normal reference rule, the least squares cross-validation procedure, two types of plug-in procedures, and a bootstrap based method. The methods are first shown to work from a theoretical point of view. A simulation study is then carried out to assess the finite sample behavior of these five bandwidth selectors. The use of the various practical bandwidth selectors are illustrated by means of data regarding the luminosity of quasars in astronomy.

• metadata
• relations

Document type	Article de périodique (Journal article) – Article de recherche
Access type	Accès interdit
Publication date	2013
Language	Anglais
Journal information	"Computational Statistics & Data Analysis" - Vol. 61, p. 107-123 (2013)
Peer reviewed	yes
Publisher	Elsevier BV ((Netherlands) Amsterdam)
issn	0167-9473
e-issn	1872-7352
Publication status	Publié
Affiliation	UCL - SSH/LIDAM/ISBA - Institut de Statistique, Biostatistique et Sciences Actuarielles
Keywords	Bandwidth selection ; Bootstrap ; Cross-validation ; Double truncation ; Kernel density estimation ; Normal reference rule ; Plug-in
Links	http://hdl.handle.net/2078.1/128768[Handle] https://doi.org/10.1016/j.csda.2012.11.017[DOI]

See also
	[boreal:110124]	Bandwidth Selection for Kernel Density Estimation with Doubly Truncated Data