Evaluation of measurement uncertainty by Bayesian-entropic approach

This paper discusses the possibility of using Bayesian inference methods and entropy optimization principles for the evaluation of uncertainty in a measurement process. In particular, the proposed approach is applied to the recalibration of a measurement device which incorporates a calibrated reference quantity. An a priori estimate and an uncertainty for the reference quantity are supposed to be available from a previous calibration. The adopted method is shown to produce a posteriori estimates for the reference quantity, regarded as the recalibrated value and for the measurand which are correlated.