In stratified k-fold cross-validation for a classification problem, what makes it 'stratified'?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because stratified k-fold ensures that each fold reflects the class distribution of the overall dataset, which is especially important for imbalanced targets to avoid folds that have no minority-class examples. A describes sorting by predicted probability, which is unrelated to stratification.
Full explanation below image
Full Explanation
B is correct because stratified k-fold ensures that each fold reflects the class distribution of the overall dataset, which is especially important for imbalanced targets to avoid folds that have no minority-class examples. A describes sorting by predicted probability, which is unrelated to stratification. C describes clustering-based splitting, not stratified cross-validation. D describes sampling, not the stratification mechanism in cross-validation.