{"blocks": [{"key": "24deaee9", "text": "Scenario", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "8abb029d", "text": "Technical deep dive – building a production model for credit-risk scoring at Capital One.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "df6f959c", "text": "Question", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "02c08919", "text": "Compare logistic regression, random forest, and gradient boosting for credit-risk modeling; discuss pros and cons. Explain how you would evaluate model performance, handle class imbalance, and ensure model interpretability.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "3f872887", "text": "Hints", "type": "header-two", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}, {"key": "63d693e0", "text": "ROC-AUC, KS, SMOTE/weighting, SHAP, compliance requirements.", "type": "unstyled", "depth": 0, "inlineStyleRanges": [], "entityRanges": [], "data": {}}], "entityMap": {}}