Prediction of CardioVascular Disease (CVD) using Ensemble Learning Algorithms

Abstract

Over the last decade, CardioVascular Diseases (CVD) and allied heart disorders have been the leading cause of death world wide. Early prediction of CVD can help high-risk patients make lifestyle changes and as a result can reduce complications. Researchers in the past have worked on developing computational models to aid health care professionals in the prediction of CVD. Most of the existing techniques lack precise feature sets and suffer from high overfitting and low accuracy. To present more accurate model of predicting CVD along with employing better feature set, ensemble learning models along with individual classification techniques are proposed. Extensive performance analyses on Kaggle Cleveland Heart Disease dataset clearly show our model can significantly improve on the accuracy and F1-score than some of the existing competitors.