ARCHIVES

Heart disease prediction is significant problem in medical research. In the present context as per the Atlas statistics around 15.5% of the world population deaths are caused due to heart diseases. The heart disease prediction can be done by analysing various factors like Age, sex, cholesterol and the like. In the present work, Cleveland heart disease dataset with 76 features was taken from UCI repository for predicting the presence of Coronary heart disease. Good features hugely impact the performance of model. We applied feature selection on machine learning algorithms: Decision tree, Support vector machine and the like combined with bagging model to improve the model accuracy .The developed model shows the output as presence or absence of Coronary heart disease. Model performance is evaluated with metrics: Accuracy score, Precision, Recall, F1-score and Jaccard index.