This article discusses maximum likelihood and Bayesian inference, particularly empirical and hierarchical Bayesian approaches, under different assumptions of the piecewise exponential (PEXP) model. PEXP assumes independent time between failures, distributed exponentially, with a mean that can experience a discrete increase or decrease between failures. It can be used to model repairable systems or systems that suffer damage on a failure. The PEXP approach is applied to a real data set using the hierarchical Bayes approach, the classic empirical Bayes approach and the parametric empirical Bayes approach.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados