A validation fold contains no positives and ROC-AUC is undefined. Which split change helps, and what limitation remains?
Classic MLLevel: IntermediateMissing class in a fold
A validation fold contains no positives and ROC-AUC is undefined. Which split change helps, and what limitation remains?
Hints
- Count minority samples before choosing folds.
- A splitter cannot create evidence.
Answer
Stratification helps retain class representation when enough examples exist. It cannot manufacture minority examples; too many folds or grouped data may still make the metric unestimable.
Source and licence: scikit-learn 1.7.2 — cross validation · BSD-3-Clause · answer: Quantoros editorial