• Manish Kulkarni
• Atharva Joshi
• Manas Kulkarni
• Arnav Kulkarni
• Prema Kadam

: As diabetes mellitus evolves into a pressing global health issue, the need for effective early detection tools is critical. Machine learning (ML) offers powerful predictive capabilities, but its use in clinical settings is often hindered by a lack of transparency in so called "black box" models. To address this, we propose a framework that integrates high accuracy ML classifiers with Explainable AI (XAI) to create both powerful and interpretable prediction systems. This study combines a literature review with an empirical case study, comparing models like Support Vector Machines (SVM), Random Forest, and CatBoost. Our findings consistently identify plasma glucose, BMI, and age as the most critical predictive biomarkers. We demonstrate how XAI techniques specifically SHAP for global analysis and LIME for individual patient explanations can demystify complex models. This approach helps close the gap between AI's predictive potential and the need for trustworthy clinical decision support, paving the way for more reliable integration of AI in healthcare.

Explainable AI (XAI), Machine Learning, Diabetes Prediction, Clinical Decision Support, Model Interpretability, SHAP, LIME, Ensemble Methods.

"EXPLAINABLE MACHINE LEARNING FRAMEWORK FOR EARLY STAGE DIABETES PREDICTION: A SYNTHESIS OF MODEL PERFORMANCE AND INTERPRETABILITY", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 10, page no.c753-c758, October-2025, Available :http://www.jetir.org/papers/JETIR2510294.pdf

"EXPLAINABLE MACHINE LEARNING FRAMEWORK FOR EARLY STAGE DIABETES PREDICTION: A SYNTHESIS OF MODEL PERFORMANCE AND INTERPRETABILITY", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 10, page no. ppc753-c758, October-2025, Available at : http://www.jetir.org/papers/JETIR2510294.pdf