Construction of a multi-connected object model object based on joint use of data and expert evaluations

The article discusses a method for constructing statistical models of multi-connected objects in the form of multidimensional regression dependencies based on the joint use of measurement data and expert assessments. The main idea of the method is to convert independently obtained expert estimates and measurement data to a single type - the law of distribution of a continuous random variable. This makes it possible to form a correlation table based on the solution of the inverse problem of determining the distribution law of the function of a random argument. The formed correlation table serves as the basis for the construction of the regression dependence by the known methods of regression analysis. The main result of the work is the first proposed method for constructing multidimensional regressive dependencies by sharing expert estimates and measurement data.