Prediction of Cardiovascular Disease Using Machine Learning

Abstract

Cardiovascular disease (CVD) poses a significant threat to human health by impairing the functionality of the heart and blood vessels, often resulting in death or physical paralysis. Early and automated detection of CVD is crucial for saving lives. While numerous efforts have been made towards this goal, there remains scope for enhancing performance and reliability. This study contributes to this ongoing endeavour by employing two robust machine learning techniques, multilayer perceptron (MLP) and K-nearest neighbour (K-NN), for CVD detection, utilising publicly available data from the University of California Irvine repository. The models’ performances are optimally enhanced by removing outliers and attributes with null values. Experimental results showcase a superior accuracy of 82.47% and an area-under-the-curve value of 86.41% achieved by the MLP model, outperforming the K-NN model. Consequently, the proposed MLP model is recommended for automatic CVD detection. Furthermore, the methodology presented herein holds promise for detecting other diseases, and the performance of the proposed model can be validated across additional standard datasets.