A Detailed Analysis of Machine Learning Algorithm Performance in Heart Disease Prediction

Human deaths are natural events, but heart disease remains a leading cause of mortality worldwide. Predicting heart disease using machine learning can significantly benefit global health, though it presents clinical challenges requiring high accuracy and efficiency. Many countries face a shortage of cardiovascular experts and a high rate of misdiagnoses, which can be mitigated by accurately determining the stage of coronary disease and supporting medical decisions with automated patient records. This paper addresses the prediction of coronary disease based on input attributes using data mining techniques. The results showed that the CatBoost Classifier achieved the highest accuracy at 87.2%, followed closely by the Ridge Classifier at 86.3%. Other models, such as XGBoost and Gradient Boosting, performed similarly, achieving 84.0%, while the RandomForest Classifier had a slightly better performance with an accuracy of 85.6%.