Polylinear regression analysis in problems of radiochemistry

Some radiochemical problems are formulated in the form of a problem of polylinear regression analysis that enables to use standard software to solve them. Application peculiarities of regression analysis to estimate contribution of different sources of the atmosphere contamination, to investigate into the irradiated nuclear fuel, to estimate concentrations using spectroscopy data, to measure neutron fields of nuclear reactor, to estimate parameters of crystalline lattices using diffraction pictures, to interpret data of X-ray fluorescent analysis, to estimate constants of complex formation, to analyze the results of radiometric measurements. Incomplete models are used to estimate only a part of the required parameters. In this case, account should be made for possible mixing of estimations. Mixing estimation algorithm realized by means of regression analysis standard programs is presented. 9 refs