APPLICATION OF THE METHOD OF GROUP ACCOUNTING OF ARGUMENTS FOR THE ANALYSIS OF PETROPHYSICAL DATA

The paper considers a range of tasks related to the processing and analysis of petrophysical data, which are effectively solved by the method of group accounting of arguments (MGUA). MGUA is a machine learning method that is an alternative to regression analysis and neural network modeling. The method was tested when working with information from the petrophysical database of crystalline and sedimentary rocks of the Voronezh Crystal Massif (VCM). The basis of the analysis technology is the formation of complex identification equations that allow you to generalize, analyze and effectively use petrophysical data. Previously, the use of such equations for estimating the ore content of nickel-bearing intrusions of the mamonsky complex was considered. Examples of solving problems of robust estimation and identification of petrophysical data are provided. It is shown that the identification models of the MGUA allow automating the procedure for detecting outliers in the data, assess the belonging of rocks to a certain material complex. Since petrophysical information is a necessary link in the geological interpretation of geophysical observations, the experience of obtaining complex models linking the density of metamorphic and intrusive rocks of the VCM with electrical resistivity and magnetic susceptibility is interesting.