Software Reliability: B ayesian Analysis
Michael Peter Wiper · Wiley StatsRef: Statistics Reference Online · 2014
Abstract In this article, the use of Bayesian methods in software reliability predictions is reviewed. In particular, the article studies software reliability models for interfailure times or for the numbers of failures in a given time period. After a general introduction to the use of the Bayesian approach to inference in the software reliability field, in particular, commenting on the use of the prequential likelihood for model checking, the article concentrates on reviewing Bayesian analyses of some of the most well known software reliability models, outlining the concatenated failure rate model of Jelinski and Moranda. The article ends with a real data example where the goodness of fit and reliability predictions of various models are compared.