فهرست نشریات علمی دسترسی آزاد ایران

This study proposes a multi-objective optimization model for predicting and managing Type 1 Diabetes by integrating clinical, behavioral, and lifestyle factors into a mathematical framework. Utilizing NSGA-II and MOPSO algorithms, the model simultaneously minimizes blood glucose deviation, insulin cost, and the complexity of input features. Synthetic patient data, modeled on clinical norms, were used to simulate diverse patient scenarios. The results demonstrate that NSGA-II provides superior convergence, diversity, and predictive accuracy compared to MOPSO, although MOPSO offers faster execution. Feature importance analysis revealed carbohydrate intake, stress, and physical activity as the most influential factors in glycemic control. The proposed model strikes a balance between predictive performance and interpretability, serving as a robust decision-support tool in personalized diabetes management. Its flexibility and transparency make it suitable for integration into clinical practice and future intelligent health systems.