Early prediction of heart disease risk using extreme gradient boosting: a data-driven analysis

Heart disease is a leading cause of morbidity and mortality worldwide. Early identification of heart disease risk is critical for timely treatment and prevention of further complications. This study provides a detailed examination of a novel heart disease dataset encompassing 333 cases and 21 features. The study employed the eXtreme gradient boosting (XGBoost) algorithm to develop an intelligent predictive model to detect the likelihood of heart disease at an early stage. The choice of the XGBoost model for this study was apt, considering its strengths in managing structured medical datasets with multiple features, resistance to overfitting, and interpretability for insights into feature importance. Feature selection was utilised to identify the most important predictors for prediction. The findings demonstrate that the Gradient Boosting classifier outperforms other machine learning (ML) techniques with a 99% accuracy rate. The results highlight the capability of ML in aiding the early detection of heart diseases.