What is k-fold cross-validation?

Updated May 17, 2026

Short answer

It splits data into k folds and evaluates model k times.

Deep explanation

Each fold is used once as test set while others form training set.

Real-world example

Used in small datasets to maximize training usage.

Common mistakes

  • Data leakage across folds.

Follow-up questions

  • What is stratified CV?
  • Why k=5 or 10?

More Model Evaluation interview questions

View all →