Transformation of types of signs for a task of the regression analysis

Today there are a large number of methods and analysis algorithms of data for solution of problems of image identification, automatic classification, component, correlation and regression analysis which in case of availability of data of the different type nature become inapplicable because they are intended only for processing of quantitative information. The approach to solution of this problem named conversion of types of signs for a problem of regression analysis is described in the offered article. At the same time conversion of type of signs is presented in the form of the separate task allowing to pass from not quantitative signs to quantitative ones and in further processing to use all range of traditional methods of the analysis of an original information. The offered algorithm is implemented in integrated environment of working out of the of Delphi 10 Seattle software.