Why should preprocessing be fitted inside cross-validation folds?
- A Fitting on the full dataset leaks information from the validation data into the transformation
- B It is computationally faster
- C It reduces memory use
- D It is required by scikit-learn
Answer
Fitting on the full dataset leaks information from the validation data into the transformation
Using a Pipeline object handles this automatically and eliminates the error structurally rather than relying on discipline.





