Resumen de A Bayesian Hierarchical Power Law Process Model for Multiple Repairable Systems with an Application to Supercomputer Reliability

Kenneth J. Ryan, Michael S. Hamada, C. Shane Reese

• Hardware failure data from the Blue Mountain supercomputer are analyzed using nonhomogenous Poisson process models fitted using Markov chain Monte Carlo methods within a hierarchical Bayesian framework. A variety of simulations are used to compare time-truncated and failure-count designs and to evaluate hierarchical modeling of multiple repairable systems as compared to separate modeling of such systems.