Machine learning has emerged as a transformative tool in healthcare, enabling predictive analytics for disease progression, patient management, and clinical decision-making. This study integrates three critical areas: mortality trends in the USA, heart failure survival prediction using machine learning (ML) models, and hospital readmission forecasting with artificial intelligence (AI)-driven methodologies. Using datasets from national health statistics, clinical trial data, and electronic health records, this research applies Logistic Regression, Random Forest, Support Vector Machines (SVM), Neural Networks, and Gradient Boosting models to enhance prediction accuracy. Results indicate that SVM achieves the highest predictive accuracy for heart failure survival (88.41%), while Gradient Boosting performs best for readmission prediction. Findings highlight ML’s potential in improving risk stratification, resource allocation, and targeted interventions, contributing to a growing body of AI applications in healthcare analytics. This study provides a foundation for future research on personalized medicine and predictive healthcare models, with broader implications for disease prevention and healthcare efficiency.