A Comparative Analysis of Machine Learning Classifiers for Robust Heart Disease Prediction

Abstract

In the past decade, Cardiovascular Diseases (CVD) have become a significant public health concern across the nations due to the treatment cost. Therefore, the low-cost cardiac health monitoring system is highly essential, which analyzes the collected data and makes robust predictions. The statistical data analysis has limitations to infer the knowledge from several parameters. On the contrary, Machine Learning (ML) models deal with numerous parameters and can successfully extract the patterns. In this paper, ML classifiers are designed, and comparative analysis is carried out for the robust heart disease prediction. Five ML classifiers such as Logistic Regression (LR), Naive Bayes (NB), Random Forest (RF), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN) are implemented, and their performance is extensively evaluated on the Cleveland Heart Disease Data set. The comparative analysis across five performance metrics confirms the applicability of ML classifiers for heart disease predictions. © 2020 IEEE.