Abstract: The article is devoted to the use of the risk function for modeling economic, in particular, production and technological
systems. Initially, the risk function was developed to assess the risks of failures in the functioning of various technical facilities in the
event of a possible occurrence of man-made cataclysms (disasters). The dependent variable was a certain generalized risk, and the
independent ones were its local, particular characteristics of risk. A decrease in any of the local characteristics of the risk cannot lead
to a decrease in the generalized risk. This paper proposes an approach that allows using a positive production indicator as a
dependent variable. In addition, some new criteria for the adequacy of risk models are proposed: a vector of “positives”, a relative
matrix of “positives”. The developed approach was applied to simulate the volume of loading of the main types of cargo by rail;
competing types of transportation in relation to rail transportation were chosen as independent variables. These are factors such as
road transport, sea transport, pipeline transport, inland waterway transport. The model was obtained as a result of solving a partial
Boolean linear programming problem using an appropriate software package. The analysis of the resulting model was carried out.
The low average relative error of approximation allows us to conclude about its good approximation characteristics. We also
calculated the contributions of the variables in percentage terms, in the form of a vector of "hits", in the form of a relative matrix of
"hits". These indicators made it possible to determine (in a model sense) the main negative factor - transportation by pipeline. Based
on the results of the analysis of the model, it can be concluded that the developed approach to applying the risk function can be
recognized as effective.
Index terms: regression, risk function, least modulus, partial boolean linear programming.