Development of models, algorithms, and software for forecasting the population of the region

The work is devoted to population forecasting using a genetic algorithm and the construction of a Lotka-Volterra mathematical model. A feature of the study is the development of a methodological approach to regional population forecasting, which is an alternative use of the genetic algorithm and the Lotka-Volterra mathematical model. This approach allows, firstly, to improve the accuracy of the forecast, and secondly, to expand the possibilities for correcting information models of regional demographic data. Variants of the genetic algorithm have been implemented: with the removal of extra individuals and without removal, and systems of two differential equations have been constructed where, as variables, fertility, mortality and migration indicators are presented in pairs. A forecast has been developed for the values of birth, death and migration indicators for the Republic of Buryatia and its regions for three and five year periods. To assess the accuracy of algorithms and predictive values, it is proposed to use the value of the difference between the integral from the predicted and from the real function, which will make it possible to determine the presence of sharp jumps in the calculated data.