Using desirability functions when making managerial decisions in the mineral resource complex

Relevance. The ever-growing flows of information, tight deadlines for decision-making, the need to maintain competitive positions require ranking calculations to identify leaders and promote leadership factors, ranking units of business entities according to selected criteria, etc. Reliability, visualization, and often acceleration of obtaining results predetermine the appeal to mathematical techniques. Purpose of the study: substantiation of choice of the most acceptable option of the desirability function for use in the process of making managerial decisions related to the development of subsoil resources. Research method: comparative analysis of options for calculating the desirability function, comparison method, methods of mathematical statistics. Results. Two options for calculating the desirability function are considered. The first one developed by E. Harrington, who proposed a special verbal-numerical scale, which made it possible to formalize the system of preferences of different experts. The partial function of desirability, in this case, is constructed so that it is close to "satisfactory" linear one. According to recommendations, the lower and upper boundaries of “satisfactorily” correspond to the minimum and maximum values of indicators for the available data array. There is a slightly different approach to the determination of the partial function of desirability described in a number of works. The generalized desirability function is a convolution of particular functions and is defined as the geometric mean or logarithmic mean. It is possible to use weighting factors with different significance of particular desirability functions. Comparison of various approaches is made for the conditions of management of logistics and equipment of OOO Gazpromtransgaz-Yekaterinburg. An analysis of the results shows that despite discreteness of the particular desirability functions, which in the first case lie in the range from 0.37 to 0.69, and give the entire range of values from 0 to 1 in the second one, the ratings for the generalized desirability function completely coincide with regard to determining a rating of quarterly values of industrial and economic activity. It follows that both methods of calculating partial functions are legitimate, however, the first option seems more convenient technically due to the absence of values of partial desirability functions close to zero, which complicate calculations using computer technology. A third version of the rating was also tested using calculated percentage estimates of the indicators under consideration with respect to the maximum desired value (taking into account their conversion into the relative indices). As in previous cases, the rating of countries according to the generalized function of desirability turned out to be comparable with previously calculated. Conclusions. Comparative calculations confirmed the validity of both options for finding particular desirability functions, since the ratings for the generalized desirability function coincide. The first version of the calculation is more convenient due to the lack of values of particular desirability functions close to zero, which simplifies the use of computer technology