Methodology for the projection of population pyramids based on Monte Carlo simulation and genetic algorithms

The analysis of the evolution of population pyramids is crucial to study and tackle the growing issue associated to the depopulation of different regions around the world. This task has been based on the application of Monte Carlo simulation for the projection of the new population from the original one using birth, mortality, immigration and emigration data to model it. It has been applied both with current trends and with modified ones according to the user selections in order to obtain and analyse possible alternative scenarios. An inverse process is also presented, where a desired population pyramid is provided, and the system, based on genetic algorithms, determines the modifications of the previously stated rates to get the most similar output. Both tools are applied to a particular situation, the municipalities of a province in the north of Spain, so the functionality of the whole methodology is shown in an empiric way. These results show that projections are accurate with respect to the real data available to contrast, as well as that the inverse system generates close pyramids to the desired ones under reasonable circumstances. Additionally, these approaches have novel functionalities and make use of techniques not widely applied in the field, thus making them innovative with respect to other noteworthy examples in the field.