Методика прогнозування кількості інвалідів з числа санітарних втрат

The article considers the problems of building mathematical models for predicting the contingent of pension recipients under conditions of uncertainty caused by the impact of hostilities. Based on the study of statistical information on the structure and dynamics of sanitary losses, an approach to revealing systemic uncertainty in the problem of predicting the contingent of pension recipients is proposed. This work is part of the study of the application of methods of intelligent data analysis and mathematical modeling in information technology intended for use in the pension system.The problem of predicting human losses as a result of hostilities is an urgent problem even in the conditions of the use of high-tech weapons. Sanitary and irretrievable losses are not only the effectiveness of the combat use of the unit, but also the costs of treatment, rehabilitation, pension provision, insurance payments. A high-intensity war with the use of the most modern weapons and military equipment has no analogues in retrospect. Therefore, approaches to predicting combat medical and non-recoverable losses, which are based solely on the calculation of average values or on analogies, cannot provide high-quality results. The study is devoted to the development of a methodology for predictive modeling of medical losses, the basis of which is probabilistic-statistical models in the form of Bayesian networks, the method of time series similarity and cluster analysis. If necessary, the proposed methodology can be used to perform calculations under different scenarios.During the study, a number of numerical experiments were conducted, in which the correctness of the application of the proposed methodology was investigated. Acceptable forecasting results were obtained.The proposals presented in the work will allow to increase the sustainability of the pension system of Ukraine, including by more accurately determining the dynamics of the contingent of pension recipients, and, accordingly, the costs of paying pensions.