Aprendizado de máquina e análise de redes sociais em grafos: um estudo do movimento dos estudantes de medicina no Brasil

Abstract

Brazil’s healthcare system has been facing challenges for decades due to the uneven distribution of doctors across the country. Few studies have attempted to address medical mobility to understand the decision-making factors that determine where professionals choose to settle. Understanding the circulation patterns of doctors in Brazil can be highly valuable for the government, as it provides insights that may lead to better job opportunity policies and help define optimal locations for new medical schools. More specifically, it is crucial to understand how medical students decide where to pursue their degrees, as this choice will influence the future mobility of professionals. This study is part of an investigation into the flow of doctors in Brazil, considering data provided by the Ministry of Health and other Brazilian research and governmental agencies. The proposed study employs machine learning techniques to derive and analyze patterns in where individuals graduate and practice medicine. Additionally, Social Network Analysis is used to assess the major migration flows of medical professionals across regions. The results indicate that states with higher centrality in the network tend to attract more doctors, both during their education and in their professional practice. Furthermore, we identified correlations between the location of medical education and the choice of workplace, highlighting that medical mobility follows concentrated patterns in specific regions. These findings can contribute to incentive policies that promote a more equitable redistribution of doctors throughout the country.