PREDICTION OF HEART DISEASE USING CLASSIFICATION AND FEATURE SELECTION METHODS

Heart disease, known as cardiovascular disease has been one of the main causes of death worldwide in recent years. It is affected by various risk factors such as high blood pressure, high cholesterol, diabetes, smoking, obesity, physical inactivity, etc. In this study, Cleveland database of UCI repository of heart disease patients is used. The dataset includes 303 instances and 76 variables which of 14 variables are considered. At first, three different feature selection techniques, namely LASSO, Elastic Net and Adaptive LASSO under specific tuning parameters determined by cross-validation, are applied to reduce the number of independent variables used for the prediction. Then, five different classification techniques, K-nearest neighbor, Support Vector Machine, Decision Tree, Random Forest, and Logistic Regression are performed to predict heart disease. Performances of these methods are measured by different metrics, such as accuracy, sensitivity, specificity, and Matthew correlation coefficient.