Coupling Bayesian Inference and Monte Carlo Methods in Error Propagation

摘要

In this article, we show how the coupling of Bayesian inference and Monte Carlo methods provides a powerful and flexible way to investigate the variation of a nonlinear transfer function, which is a function of both ran- dom variables and parameters. The proposed coupling naturally accounts for the uncertainty in the parameters which are estimated from data and is illustrated with the assessment of buoyancy-corrected weights from a weighing mea- surement system.