Diseño de algoritmo compuesto por Machine Learning y un modelo probabilístico para la detección de diabetes

Abstract

Diabetes mellitus (DM) is a type of metabolic disorder which causes chronic hyperglycemia. This alteration usually occurs due to an inadequate secretion of insulin. In the present work, a set of algorithms for the detection and prediction of diabetes was carried out using Pimas database. The algorithms used were: a Naive Bayesian algorithm with an 79.67% F1, a KNN algorithm with an 79.64% F1, an Artificial Neural Network (ANN) with an 74.07% F1 and an algorithm composed of the three previous algorithms with an 80.32% F1.