Heart Disease Prediction using Feature-Optimized ML and Hybrid Deep Learning Models

Cardio disease remains as foremost reason of global fatality, emphasizing the need for accurate and early prediction systems. This study proposes an integrated framework that combines feature-optimized deep_learning (DL) and machine_learning models to classify heart diseases. Principal Component Analysis (PCA) and Relief feature selection methods were chosen to decrease the dimensions and extract relevant attributes. The refined dataset was used to train high-performing classifiers including XGBoost, SVM, and Random Forest. Additionally, a hybrid DL model combining DenseNet and residual connections was developed to collect intricate form in the UCI data repository. Evaluated on UCI dataset for Heart disease, the proposed framework accomplished a classification accuracy of 96.4%, surpassing existing benchmarks. This comprehensive approach offers a reliable and interpretable result for proactive identification of cardio disease. The results demonstrate the effectiveness of integrating feature optimization and hybrid modeling in predictive healthcare systems.