Asymptotic normality for estimators of the additive regression components under random censorship

摘要

We establish asymptotic normality for estimators of the additive regressioncomponents under random censorship. To build our estimators, we couple themarginal integration method (Newey (1994)) with an initial Inverse Probabilityof Censoring Weighted estimator of the multivariate censored regressionfunction introduced by Carbonez et al. (1995) and Kohler et al. (2002).Asymptotic confidence bands are derived from our result.