Modelowanie salda migracji dla miasta Poznania z wykorzystaniem sztucznych sieci neuronowych
Modeling the migration balance for the city of Poznań with the use of artificial neural networks

Summary/Abstract: In the article two goals were achieved. The first goal was to model the internal migration balance for the city of Poznań depending on selected economic and social factors (i.e. the number and area of existing apartments, the number of rooms in an apartment, the average usable floor space of an apartment per 1 person, internal migration balance, population change per 1000 inhabitants, total population, divided into women and men, and population density per square kilometer) in 2005–2020 based on statistical data from the Central Statistical Office. The research area covered 22 communes, which together with the city of Poznań formed the Urban Functional Area. Another research goal was to develop an experimental model of an artificial neural network to predict the migration balance for the city of Poznań at two time points – 1 year and 2 years ahead. It is worth noting that artificial neural networks were used in the study, which is a methodological approach that is relatively rarely used in the literature for this purpose.This paper showed that it is possible to use artificial neural networks to accurately predict the migration balance for the city of Poznań. The research showed that the analyzed economic and social factors for the communes included in the Urban Functional Area of Poznań were significantly correlated with the migration balance of the population for the city of Poznań, and thus can be used to predict the migration balance for the city of Poznań. The obtained results confirm the maintenance of a negative migration balance for the city of Poznań with a decreasing trend and a positive migration balance in the neighboring communes with a decreasing tendency.

• Details
• Contents