Automated prediction of Coronary Artery Disease using Random Forest and Naive Bayes

In recent days, the world is suffering from Coronary Heart Diseases. It has become a widely spread disease. CAD have a strong side effects, which might cause physical symptoms and family burdens. Other physical symptoms might increase the patients' rate of death in the general population. Therefore, the study focused on developing predictive model for the diagnosis of CAD. The model was developed using computational machine learning techniques that incorporate Random Forest (RF), and Naive Bayes (NB) techniques. The study will use an open source dataset named Z-Alizadeh sani, which will assist in the diagnostic process. Naive Bayes outperformed the other studies in literature review with 100% sensitivity and Negative predictive rate of 100%. Naive Bayes outperformed when compared with Random Forest with an accuracy of 83% with 13 features. The achieved results are encouraging to use the built models as supportive software in the diagnosing process.