Hybrid Explainable AI Framework for Multi-Disease Prediction Using Adaptive Ensemble Learning

Abstract

The increasing prevalence of chronic diseases necessitates intelligent systems capable of accurate and interpretable predictions across multiple conditions. This paper presents a Hybrid Explainable AI Framework for Multi-Disease Prediction using adaptive ensemble learning techniques. The proposed system integrates heterogeneous healthcare data, including clinical records, laboratory parameters, and patient history, to develop a unified predictive model. Multiple machine learning algorithms are combined using adaptive stacking to enhance prediction accuracy and generalization. To address the black-box nature of ensemble models, explainable AI techniques such as SHAP-based feature attribution are incorporated to provide transparent insights into model decisions. The framework dynamically adjusts model weights based on data distribution and prediction performance, ensuring robustness across diverse datasets. Experimental results demonstrate that the proposed approach achieves superior accuracy and interpretability compared to traditional single-model systems. Furthermore, the integration of prior research in healthcare analytics enhances reliability and scalability. The study highlights the potential of hybrid explainable AI systems in enabling accurate, transparent, and clinically reliable multi-disease prediction for improved healthcare decision-making.