PROCESSING OF ISCHEMIC HEART DISEASE DATA USING ENSEMBLE CLASSIFICATION METHODS OF MACHINE LEARNING

Abstract

The WHO 2019 statistics provide evidence that cardiovascular diseases are among the prevailing causes of death globally [1]. In this study, a combined dataset of coronary artery disease (CAD), also known as ischemic heart disease, was used as the dataset for analysis. To influence the outcome of the occurrence of cardiovascular diseases, it is important to find significant features that contribute to the presence of this disease. This article demonstrated that important features can be obtained through classification and their visualization in Tableau. Three classification models were built, and important features were identified for each model. Then, the top 10 important features were selected from each model, and through comparison, the 5 most important features were identified that may influence the disease outcome. The classification models achieved the following f1-score results: LGBM (93.2%), XGB (92.0%), and RF (89.1%).