RESEARCH OF APPROACHES TO CONSTRUCTING PREDICTIVE MODELS OF THE SPREAD OF VIRAL DISEASES

The article is devoted to the analysis of the available mathematical models in epidemiology and the possibility of their modification. We note that the situation with the COVID-19 virus pandemic is characterized by several features not comprehensively studied in the existing models. For a rational response to existing challenges, it is necessary to have a predictive and analytical apparatus in the complex (national and regional scale) mathematical models with a planning horizon of 2 years (the expected period of mass production of vaccines). The article discusses the existing approaches to predicting the spread of the COVID-19 virus in Russia based on mathematical models of epidemics. The possibilities and limitations of the proposed approaches are considered. In the conditions of the Russian Federation, transport connectivity at the interregional and intraregional levels plays an important role, and for megalopolises - transport flows within large agglomerations and the age structure of the population. In contrast to previous pandemics and epidemics, public policy plays a significant role. The approach, which consist in building multi-agent models that combine the advantages of compartment models and models based on the Monte Carlo method (individually oriented) is proposed by the authors. It is planned to use compartment models to assess the dynamics of the process and individually-oriented models - at the level of individual territories and districts.