Posterior Distribution and Loss Functions for Parameter Estimation in Weibull Processes

Abstract

Counting processes are used in modelling of repairable systems. In Bayesian parameter inference the choice of fitted prior distributions and loss functions has great importance. This paper deals with studies of prior and posterior densities. Simulation enables comparisons of estimators obtained with different loss functions. Especially, in the case of a Weibull process, properties relative to prior and posterior distributions are investigated.