Heart Failure Mortality Prediction: A Comparative Study of Predictive Modeling Approaches

This study presents a comparative assessment of various machine learning models for predicting mortality in heart failure patients. Through a rigorous analytical approach, we have scrutinized models ranging from logistic regression and support vector machines (SVM) to advanced ensemble algorithms like Random Forest and XGBoost. Our analysis delves into the accuracy, sensitivity, and specificity of each model, utilizing real clinical data to validate our predictions. The results indicate that while traditional models such as logistic regression maintain robust performance, it is the ensemble algorithms that stand out for their superior predictive capability, evidenced by areas under the curve (AUC) close to 0.90. The findings underscore the transformative potential of machine learning techniques in the prognosis and management of heart failure, providing crucial insights for early intervention and improving clinical outcomes in high-risk patients.