We discuss time series cross-validation (expanding and sliding window), evaluation metrics, baselines, and practical guidance.
