Identification of cardiovascular disorders using machine learning classification algorithms

Abstract

Early detection of myocardial infarction is crucial for necessary medical support and reducing its mortality rate. Every year a huge amount of people are suffering and dying of different heart diseases. The advent of Machine Learning techniques to learn and predict future events based on the data has brought about revolutionary changes in the field of healthcare. These techniques can be used to predict heart disease, and also to identify the type of disease that the patient is suffering from. In this work, we have used a dataset that contains the clinical records of patients who have been admitted into a hospital with a heart problem and experimented with different classification algorithms to predict the type of heart problem that the patient got. We have experimented with the dataset from a different perspectives and a thorough discussion reveals that XGB ensemble classification performs best for this multi-class classification problem. This algorithm gives the best evaluation metric of 99% balanced accuracy, 0.99 ROC AUC, and a perfect F1 score.