Statistical Indicators for the Optimal Prediction of Failure Times of Reliability Stochastic Systems: A Rational Expectations-Like Approach

The estimation of the optimal failure time of stochastic systems is the aim of this paper. We are in the field of reliability theory, and we merge the Bayesian-like method of estimating the failure time of the so-called k-out-of-n systems with rational expectations-like expectations, thus obtaining a new forecasting method and innovative and original results. To do this, we simulate the path of two sets of systems from zero time to failure time, using a series of statistical measures that allow us to compare the simulation procedures and understand which is the best predictive combination.