How do you evaluate and compare different machine learning models?

Model evaluation requires multiple metrics and validation strategies to ensure robust performance assessment.\n\n• Cross-validation: K-fold, stratified, time-series splits\n• Classification metrics: Accuracy, precision, recall, F1-score, ROC-AUC\n• Regression metrics: MAE, MSE, RMSE, R²\n• Statistical tests: Paired t-test, McNemar test for significance\n\nExample: Compare models using 5-fold CV, plot learning curves, analyze confusion matrices, and use statistical tests to determine significant performance differences. Consider computational cost and interpretability requirements.