Optimizing Diabetes Diagnosis: HFM with Tree-Structured Parzen Estimator for Enhanced Predictive Performance and Interpretability