Predicting the prevalens of Alzheimer’s disease and dynamic games against nature

The article proposes a game-theoretic approach to predicting the prevalence of Alzheimer's disease in the world based on a dynamic game against nature. The dynamics of the spreаding process is described using an integral inflow-outflow model previously used by the scientific team of the Center for Analytics of Dynamic Processes and Systems of St. Petersburg State University to predict the epidemic process of the COVID-19 pandemic and predict the population of the World and some regions and countries. The results of retrospective forecasting of the dynamics of the main variables of the integral inflow-outflow model based on the construction of program strategies of the decision-maker on the rate of change in percentage increases in integral inflow and outflow volumes and taking into account the approximation of actual values of percentage increases calculated on the basis of data on the prevalence of the disease during the years preceding the construction of forecasts. Retrospective forecasts of 2005, 2010 and 2015 have MAPE errors of no more than $1\%$. Long-term prediction of Alzheimer's disease prevalence is based on the same approaches as retrospective prediction, as well as model-based forecasting.