Cardiovascular Disease Prediction using Machine Learning Methods

Computer-aided automatic diagnosis systems are used to speed up diagnosis and treatment and to help make the best decision. In this study, eight different classification methods are considered for the detection of cardiovascular disease and their performances are analyzed separately. Firstly, the preprocessing step is carried out and the original data set is normalized with the Standard Scaler. Then, the original data is used splitting into train and test with proportions of 70% and 30%, a classification step is applied by using eight different algorithms including K-Nearest Neighbor (K-NN), Bernoulli Naive Bayes (BNB), Multilayer Perceptron (MLP), Support Vector Machine (SVM), Decision Tree (DT), AdaBoost, Logistic Regression (LR) and Random Forest (RF). The success of the classifiers is observed for four different performance criteria. Accordingly, the RF classifier exhibits the best performance with %99 accuracy. In addition, it obtains precision, kappa, Mean Square Error-MSE values as 0.994, 0.980, 0.01, respectively.