HAZARD FUNCTION MODELING USING CROSS-VALIDATION - FROM DATA-COLLECTION TO MODEL SELECTION

A general methodology for reliability modeling of component failures a nd model discrimination using cross validation is developed. First, th e requirements for collection of failure, maintenance, and operation d ata are outlined, including left and right censored data. Cross valida tion is then used as a probabilistic measure of predictive performance for selection of the optimal model from a set of reliability model ca ndidates. In addition, cross validation is used to determine the class ification or hierarchical decomposition of systems into component-clas ses which provides the best overall set of predictive models. As a mea sure of predictive performance for model selection, we demonstrate tha t cross validation is superior to likelihood function maximization and modeling error minimization, since both have bias for over-parameteri zed models and may not be generally applicable to reliability models w ith wear and shock variables, and with as-good-as-old maintenance. Cas e studies are used to demonstrate these points and the overall methodo logy.