Simultaneous statistical analysis of large arrays of correlated neutron measurements

The problems due primarily to the limitations of modern computers, preventing adoption of modern methods for estimating neutron data in everyday practice and making it difficult to use the estimated data in applied programs, are discussed. Possible solutions of these problems are presented. Specifically, an identity transformation of the statistical functional to be minimized is found: this transformation makes it possible to avoid the construction, storage, and numerical inversion of superlarge covariation matrices of measurement uncertainties. The transformation makes it possible to perform statistical analysis of an arbitrarily large number of measurements.