A Comprehensive Analysis of Five Machine Learning Models for Predicting Heart Disease

Cardiovascular disease, especially heart disease, is one of the leading causes of global mortality. Annually, approximately 17.9 million people suffer from heart disease and accounting for over 80%of the deaths. So, predicting and detecting heart disease at an early stage is very important. As a result, medical professionals need to take appropriate and necessary actions at earlier stages. By applying machine learning technology, Healthcare professionals can diagnose cardiac conditions more accurately. Many researchers are focusing on developing intelligent systems that can accurately diagnose them using electronic health data, with the aid of machine learning (ML) algorithms. This study evaluates five machine learning algorithms, including Logistic Regression (LR), Naive Bayes (NB), K-Nearest Neighbours (KNN), Decision Tree (DT), and Random Forest (RF), to predict the likelihood of heart disease using the UCI Cleveland Heart Disease dataset. Data pre-processing and feature selection steps were done before building the models. The performance of this algorithm has been evaluated using accuracy, precision, recall, and F1-score. Further, the model performance has been shown through AUC and ROC curves. It is observed from the result that Logistic Regression performs better than other considered classification models with 88.52% accuracy, 90.62% recall, and 89.23% F1-score. However, in terms of precision, Naive Bayes performs better than the other considered models. These findings highlight how machine learning-based techniques can enhance the identification of early cardiovascular risk and help clinical decision-making.