CPIT Goodness-Of-Fit Tests for Reliability Growth Models
Olivier Gaudoin · Birkhäuser Boston eBooks · 1999
Assume that the n first failure times of a repairable system are observed. In order to choose an appropriate stochastic model for these data, goodness-of-fit tests have to be performed. The aim of this work is to study the goodness-of-fit tests based on the Conditional Probability Integral Transformation of O’Reilly-Quesenberry. The general CPIT methodology is described. Then the CPIT tests are derived for the Homogeneous Poisson Process, Jelinski-Moranda model, Goel-Okumoto model and the Power-Law Process. The power of these tests is assessed through simulations and finally, they are applied to real software reliability data.