Time Series Cross-Validation
A cross-validation split where training always sits strictly before the validation block it's scored on, since a shuffled fold would let a model see the future.
An expanding window grows its training set fold over fold, always validating on the block immediately after; a sliding window keeps the training length fixed and moves both windows forward together. Both enforce the same rule: no validation row is ever trained on by a row that comes after it in time.
Shuffling a time-ordered fold the ordinary way lets the model interpolate between rows on both sides of the point it's scoring, which inflates the reported score in a way no later tuning corrects.