Relative error prediction for twice censored data (Q2304847): Difference between revisions

The paper deals with the nonparametric estimation of the regression function for twice censored survival data based on the minimization of the mean squared relative error. The construction of such an estimator is based on the key results due to \textit{H. Park} and \textit{L. A. Stefanski} [Stat. Probab. Lett. 40, No. 3, 227--236 (1998; Zbl 0959.62063)] and \textit{V. Patilea} and \textit{J.-M. Rolin} [Ann. Stat. 34, No. 2, 925--938 (2006; Zbl 1092.62099)]. Also, the estimate proposed in this paper extends the work of \textit{M. C. Jones} et al. [J. Stat. Plann. Inference 138, No. 10, 2887--2898 (2008; Zbl 1140.62031)]. Under suitable assumptions, the uniform almost complete convergence over compact sets of the introduced estimator is established. Consistent estimation of the variance allows explicit highlighting of the corresponding confidence intervals. Numerical simulations prove the performance of the proposed model.