Cross-Validation

Also known as: cv

Cross-validation splits data into folds, training on some and testing on the rest in rotation, to check whether a model generalises rather than memorises. In trading it must respect time order (no shuffling) to avoid look-ahead bias and overfitting.