A hybrid machine learning & deep learning pipeline for predicting heartdisease from clinical data. Includes EDA,preprocessing with ColumnTransformer, models (Logistic Regression, Random Forest, XGBoost) and a Keras ANN. Evaluation covers ROC-AUC, Precision-Recall, calibration, with SHAP explainability ensuring transparency and trust in healthcareAI
random-forest tensorflow scikit-learn keras cnn xgboost logistic-regression model-calibration explainable-ai shap heart-disease-prediction column-transformer confusion-matrix-heatmap roc-auc-curve stratified-kfold precision-recall-f1-score
-
Updated
Sep 21, 2025 - Jupyter Notebook