Enhancing Heart Attack Risk Prediction with Ensemble Machine Learning and Feature Engineering

heart disease continues to be the principal reason behind global deaths which imposes serious healthcare system complications. An innovative approach using ensemble machine learning models with advanced feature engineering methods recommends new methods that boost heart attack risk prediction accuracy. Our model creates more reliable predictions because it unites Random Forests and SelectKBest with ANOVA score as feature engineering methods that exploit multiple algorithms' strengths. The suggested method determines heart attack risk assessments through three risk levels including low, medium, and high. The ensemble models outperform traditional single models with a superior performance that reaches 98.67% average precision, 97% recall, and 97% F1-score, across all the classes.