A NOTE ON THE PREDICTION AND TESTING OF SYSTEM RELIABILITY UNDER SHOCK MODELS

The realibility of finite systems is studied. The functioning of a system is characterized by a vector of binary variables. The a priori distribution of the random vector permits to establish a linear probability model. It is used for deriving a predictor of the realibility and its error. Under a set of mild conditions T- Student based inferences can be made when a sufficiently large sample size is uded. The independent case and a shock dependent model are studied.