Question

A data scientist trains a BigQuery ML boosted tree model daily on an orders table that grows continuously. They want the model evaluation metrics to be comparable across daily training runs, and they need each row deterministically assigned to either training or evaluation so that a given order always lands in the same split even as new rows arrive. Which approach best meets these requirements?

Accepted Answer

Use DATA_SPLIT_METHOD='CUSTOM' with a precomputed split column derived from a deterministic hash (e.g., FARM_FINGERPRINT) of the order_id

Answer

Use DATA_SPLIT_METHOD='RANDOM' with DATA_SPLIT_EVAL_FRACTION=0.2 so BigQuery ML shuffles rows on each run

Answer

Use DATA_SPLIT_METHOD='AUTO_SPLIT' and let BigQuery ML decide the split based on table size

Answer

Use DATA_SPLIT_METHOD='SEQ' on the order timestamp so the newest 20% of orders are always used for evaluation