CMStatistics 2016: Start Registration
View Submission - CMStatistics
B0888
Title: Asymptotic efficient estimation under random censorship models Authors:  Mareike van Heel - Universiteit Hasselt/ FH Aachen (Germany) [presenting]
Abstract: The aim is to prove that under the random censorship the Kaplan-Meier integral estimators $\int{\varphi(z)F^{KM}(dz)}$ are asymptotically efficient with respect to all regular estimators of $\int{\varphi(z)F(dz)}$, where $\varphi$ is an arbitrary Borel-measurable function. We calculate a lower bound for the variance of any regular, asymptotic linear estimator of $\int{\varphi(z)F(dz)}$ and show that this bound is equal to the asymptotic variance of the classical Kaplan-Meier integral estimators.