Um estudo dos municípios de Minas Gerais através da aplicação dos mapas auto-organizáveis de Kohonen sobre o IDHM

Abstract

This study analyzes the socioeconomic development of the municipalities of Minas Gerais, through the application of Kohonen Self-Organizing Maps (KAM) to the components of the Municipal Human Development Index (MHDI). It seeks to find relationships between data not captured by the MHDI, such as discrepancies between the values of education, income, and longevity. The results revealed nine distinct clusters, with clear socioeconomic profiles and a strong pattern of regional distribution: the best-performing clusters are concentrated in the Metropolitan Region of Belo Horizonte, the Triângulo Mineiro region, and Southern Minas Gerais, while the worst-performing clusters predominate in the North and Northeast of the state, especially in the Jequitinhonha and Mucuri Valleys. The study demonstrates the potential of the Kohonen Self-Organizing Map (KAM) to identify relationships between data and segment municipalities in Minas Gerais by their characteristics.