Predictive analytics for cardiovascular disease diagnosis using machine learning techniques

Abstract

Medical data can be mined for effective decision making in the presence of disease analysis. Globally, cardiovascular alias heart disease is one of the highly rated causes of death [1] disease which will lead to 76% of the deaths [2] by the year 2030. Currently, the techniques of machine learning and predictive analytics have proven importance in medical data analysis. In a nutshell, this paper aims to apply six classifiers namely artificial neural network, support vector machine, decision tree, nearest neighbor, linear discriminant analysis, random forest, and to predict the presence of heart diseases in the patient’s datasets. Moreover, the performance of these classifiers on three heart disease datasets is compared. The results reveal that RF, LDA, DT, and ANN have performed better than SVM and KNN in terms of accuracy, recall, F1-score, confusion matrix, and error rate.