Cross-validation

k-fold cross-validation splits the data into $k$ equal folds. For each fold, train on the other $k-1$ and evaluate on that one. The final score is the average over the $k$ runs (often reported with its standard deviation).

Every example is used for validation exactly once and for training $k-1$ times, so you use all the data and get a less noisy estimate than a single split. $k=5$ or $10$ are standard.

• Stratified k-fold keeps class proportions the same in every fold — use it for classification.
• Leave-one-out ($k=n$) is nearly unbiased but expensive and high-variance.
• Time-series split always trains on the past and validates on the future.
• Group k-fold keeps all examples of one patient or user in the same fold to avoid leakage.