Random Forest vs XGBoost: Mechanics and Evaluation Metrics

Question: Compare Random Forest and XGBoost - how does each model work, and how would you evaluate them? Key points: Random Forest is a bagging ensemble of independently trained decision trees (reduces variance via averaging/voting); XGBoost is a boosting ensemble that builds trees sequentially, each correcting the errors of the previous ones via gradient descent on a loss function (reduces bias, often needs more careful regularization/tuning). Evaluation was discussed via confusion matrix, precision/recall, and F1 score, especially relevant for imbalanced classification problems.