Enhancing Trust in Ensemble Machine Learning Formulations for Diabetes Prognosis: A Comparative Analysis of SHAP and LIME Architectures for Black-Box Model Interpretability. TSP 2026, 2 (2).