scientific article; zbMATH DE number 3868431

From MaRDI portal
Publication:3334794





The kernel estimator and the nearest neighbor estimator are widely used tools for estimation of a density function in the classical case. In the paper their adaptation to randomly censored data using Kaplan-Meier estimator is being considered. An influence function approach is used to study the adapted kernel estimator. Its asymptotic variance is derived and pointwise \(L^ 2\)-consistency is proved. Moreover, two different classes of estimators arising naturally in the random censorship model are proposed, and consistency and weak convergence results under appropriate conditions are proved for them. In particular for the estimator steming from the NN estimator it is shown that its variance is censoring free.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3334794)