This paper is concerned with the nonparametric estimation of regression quantiles ofuda response variable that is randomly censored. Using results on the strong uniformudconvergence rate of U-processes, we derive a global Bahadur representation for audclass of locally weighted polynomial estimators, which is sufficiently accurate forudmany further theoretical analyses including inference. Implications of our results areuddemonstrated through the study of the asymptotic properties of the average derivativeudestimator of the average gradient vector and the estimator of the componentudfunctions in censored additive quantile regression models.
展开▼
