Uma abordagem baseada em redes neurais artificiais para diagnóstico de diabetes

Abstract

Diabetes has great incidence worldwide and is responsible for the decrease in the quality of life of many people. It is known that previous diagnosis of this disease is very important, as it allows immediate care and risk prevention. The diagnosis and classification of diabetes are the subject of discussions over several decades, and the diagnosis is considered a complex task, due to the fact that the few existing methodologies are based mostly on glycemic changes in the individual’s body. Therefore, the present work proposes an approach based on artificial neural networks for the diagnosis of diabetes.