Selection of the optimal neural network for analysis of the demographic situation

Abstract

Demographic development has been a subject of study for many years, however, even during the prosperity of new technologies, there are phenomena that affect the demographic indicators of the whole world, and for which humanity is ready. Therefore, this work is devoted to the study and identification of the main factors affecting demographic processes, analysis of the advantages and disadvantages of existing tools and the actual modeling of demographic processes in Ukraine. The aim is to study the basics of forecasting demographic processes, modeling demographic processes. It provides an overview of some modern mathematical methods for describing dynamic processes, as well as methods of approximation and forecasting. The formation of a new national strategy aimed at human development, combined with unfavorable trends in the demographic reproduction of the population of Ukraine, determines the relevance of demographic research, the novelty is to consider the dynamics of demographic processes in the world, taking into account the latest events of 2020 and assess their impact on the population. Every citizen of his country influences its development. To be a global leader, a country needs a strong economy, which in turn depends on demographics. The latter play an outstanding role within the state and outside. The population size the desire to form something, more and improve the quality of the state clearly demonstrate, possible changes in the direction of the policy vector within the state. To avoid probable social, economic or other problems, it is necessary to monitor the demographic situation, namely, to carry out demographic analysis, which can further help prevent the problems mentioned above. This is often forgotten, which can lead to not very good consequences, up to a systematic decrease in the population. Therefore, it is necessary to make the right decisions in this direction to help the country develop. It is a good idea to build forecasting models that will help you generate demographic forecasts that can be used for further processing and analysis. If you define the problem in advance, you can prevent more global ones. Today there are various software applications that allow predicting processes. Everything that surrounds us is developing, especially the digital sphere, and now we have the opportunity to do more complex calculations and even more speed. The demographic processes are considered in the work, and also the basic signs influencing a demographic situation are allocated. A multilayer perceptron, which belongs to neural networks of the direct propagation type, was chosen for the work.