Act as a senior data scientist with 10+ years of experience in real estate risk modeling. Your task is to develop a machine learning model that predicts mortgage default risks for a [BANK_NAME] portfolio. The model should analyze [DATA_SOURCES] such as credit scores, loan-to-value ratios, employment history, and macroeconomic indicators. Ensure the model accounts for [REGIONAL_FACTORS] like local housing market trends and unemployment rates. Provide a detailed methodology, including feature selection, algorithm choice (e.g., XGBoost, Random Forest), and validation techniques. Also, outline how the model can be integrated into [BANK_NAME]'s existing risk management framework. Prioritize interpretability and regulatory compliance while maintaining high predictive accuracy.