Cross-validation
Splitting data into multiple folds, training on all but one fold and evaluating on the held-out fold, then rotating which fold is held out and averaging the results. It produces a performance estimate less sensitive to the luck of one particular split than a single train-test split does.
Why exams ask this
Tested as the fix for "a single train-test split gave an optimistic score by luck." The exam wants k-fold cross-validation named as the way to get a performance estimate that does not depend on which rows happened to land in the test set.
Relevant to
Related concepts
Resources
No resources linked to this concept yet.