A band-shell no-show Linear Learner and an XGBoost each miss different nights. Averaging or voting their scores beats either model alone. Which refine step is that?
Select an answer to reveal the explanation.
Short Explanation
Linear Learner and XGBoost miss different nights. Average or vote their scores and the complementary misses cancel. That is ensembling, not dropping the stronger model, not Rekognition, and not a third model sitting on the scores.
Full Explanation
Ensembling combines independently trained models by averaging or voting so complementary errors cancel. Dropping the stronger model wastes that gain. Rekognition is not an ensemble. Training a meta-learner on the base scores is stacking, which is the next item, not a flat average.