The Accuracy of Supervised Machine Learning Algorithms in Predicting Cardiovascular Disease

In recent years, the application of machine learning models is becoming more prevalent in identifying, predicting, and diagnosing cardiovascular disease (CVD). CVD is known to be the top cause of death globally and about 17.9 million lives are lost each year due to CVD, according to the World Health Organisation. Studying the accuracy of machine learning algorithms is significant in identifying a predictive model which is most efficient in diagnosing cardiovascular diseases in patients, on the basis of their health history. In this research, the CVD dataset (accessible from Kaggle), which has 70,000 records of different patients’ case histories and 11 attributes for each patient, has been used. With the aim of identifying the most efficient model, different supervised machine learning algorithms are built and compared, implementing the decision tree classifier, the random forest classifier, K-nearest neighbors, and support vector machine classifier using Python 3.7. The results showed that the support vector machine predictive model gives a better result than the other models, in terms of the accuracy, Area Under the Curve (AUC) score and F1 score. The contribution of this research will be the application of a predictive model through which CVD can be identified, diagnosed, and treated at early stages.