Diabetes Prediction using BPSO and Decision Tree Classifier

The diabetes mellitus causes different treatments and tests as well as the complications due to this disease. Even if there are prevention methods that can reduce the risk of having diabetes, sometimes it is impossible to avoid it as for type 1 diabetes. In these cases, the only solution is to be able to diagnose it very early and do everything possible to combat the complications. This paper presents a machine learning approachto help with the diagnosis using the classification of learning in the prediction of type 2 diabetes. Attributes selection is done using binary particle swarm optimization (BPSO). Selected attributes are classified by Decision Tree classifier.