Université Paris 6
Pierre et Marie Curie
Université Paris 7
Denis Diderot

CNRS U.M.R. 7599
``Probabilités et Modèles Aléatoires''

Interval censoring : a nonasymptotic point of view


Code(s) de Classification MSC:

Résumé: Our purpose in this paper is to build some very simple piecewise constant estimators for interval censoring cases I and II and to derive some of their properties, in particular some nonasymptotic risk bounds. Such risk bounds, although less precise than the asymptotic ones which are known to hold for the maximum likelihood estimator, are easier to derive and require less stringent assumptions. One can in particular easily consider the problem of a deterministic design of the observation times.

Mots Clés: Nonparametric estimation ; cumulative distribution function ; interval censoring

Date: 1999-09-21

Prépublication numéro: PMA-528

Postscript compressed file (gzip) : PMA-528.ps.gz