Analytic Approach for Heart Disease Prediction using Supervised Machine Learning Algorithms

Abstract

According to a recent study of WHO, heart linked ailments are increasing day by day. More than 17.9 million individuals get scummed every year due to this disease. With growth in population, it becomes so much problematic to analyze and start treatments at initial stages of disease. But recent technological advancements have made it possible to accelerate medical diagnosis by early prediction of health-related issues. Therefore, the prime motive of this research is formation of a ML model for heart diseases predictions dependent upon the interrelated parameters. In our research work three machine learning algorithms are implemented, Decision tree, Random Forest and Support Vector Machine (SVM). Out of implemented algorithms SVM give the best accuracy which is 82.01% and followed by decision tree 78.94%. Random forest algorithm gave the minimum accuracy. Accuracy more than 70 % is measured good, but if accuracy is very high than it may be case of over fitting. Therefore, accuracy around 80% is commendable.