Abstract: We consider the problem of nonparametrically estimating the conditional quantile function from censored dependent data. The method proposed here is based on a local linear fit using the check function approach. The asymptotic properties of the proposed estimator are established. Since the estimator is defined as a solution of a minimization problem, we also propose a numerical algorithm. We investigate the performance of the estimator for small samples through a simulation study, and we also discuss the optimal choice of the bandwidth parameters.
Key words and phrases: Censoring, kernel smoothing, local linear smoothing, mixing sequences, nonparametric regression, quantile regression, strong mixing, survival analysis.