Early heart disease prediction using ensemble learning techniques

Abstract

Abstract Cardiovascular illnesses claim the lives of 18 million individuals each year (heart-related diseases). According to the WHO, heart disease is to blame for 31% of all deaths worldwide. In this study, a new machine learning model for predicting heart disease is provided. The proposed method was evaluated on Kaggle and the University of California, Irvine datasets. We used sample approaches and feature selection methods to identify the most useful characteristics in the dataset that was unbalanced. Eventually, classifier models were employed, and an ensemble classifier generated great accuracy. In two datasets, the proposed approach showed to be accurate in predicting heart disease. In all cases, Python was used.