A Bayesian theory of measurement uncertainty

The theory of measurement uncertainty outlined forms the mathematical foundation of the results of international discussions of many years. In contrast to the usual approaches in the literature, it is based entirely on Bayesian statistics and on the principle of maximum entropy. Individual and mutual uncertainties of measurement results are expressed and treated by a covariance matrix in conformity with international recommendations. The theory is universally applicable to most measurement data evaluation tasks including complex nonlinear adjustments and, in particular, in cases where the well-established least-squares or maximum-likelihood techniques fail, as in strongly underdetermined few-channel particle energy spectrum unfolding problems.