Published Paper Details:

Healthcare datasets frequently exhibit severe class imbalance, where critical conditions represent minority classes, posing significant challenges for traditional machine learning approaches. This research investigates the development and implementation of interpretable ensemble methods using Python libraries to address class imbalance in healthcare applications while maintaining model transparency for clinical decision-making. Our proposed framework integrates advanced resampling techniques with ensemble learning algorithms, incorporating explainable AI components through SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations) to provide clinically meaningful insights. The methodology combines traditional ensemble approaches like Random Forest and Gradient Boosting with novel weighted voting mechanisms specifically designed for healthcare contexts. Through extensive experimentation on multiple healthcare datasets including diabetes prediction, heart disease diagnosis, and cancer detection, our framework demonstrates superior performance metrics . The implementation leverages core Python libraries including scikit-learn, pandas, numpy, and specialized packages like imbalanced-learn, making it accessible for healthcare data scientists. Results indicate that our interpretable ensemble approach not only addresses class imbalance effectively but also provides actionable insights that align with clinical knowledge, potentially reducing diagnostic errors and improving patient outcomes in real-world healthcare settings.