The Bayesian Kaplan Meier Model Under the Classical Nonparametric Bootstrap

Abstract

This study aims to introduce and encourage the other to use bootstrap methods in statistical survival analysis. We show how to bootstrap the Kaplan-Meier Bayesian estimator and pay attention to its advantage unlike the classical Bayesian analysis of the Kaplan Meier method. In our study, we essentially try to focus on the application of Kaplan Meier Bayesian models in the estimation of unemployment durations of those registered with the local employment agency of Ain El Benian, with the aim of to determine the role of bootstrap in improving the quality of estimation. We find that the choice between the period models of the same family is likely to produce a multitude of decisions from the results of the application. A Bayesian survival method based on the Kaplan Meier model with the classical nonparametric bootstrap provides realistic solutions for a number of individuals of different nature, simple and relatively easy to exploit numerically global durations.