Sort transforms by ownership boundary
Decide where each feature transformation should live in the ML pipeline.
Sort each transformation by where it should be owned first. Train-fit pipeline artifact Pure formula feature logic Request-time online feature Monitor-only validation Median value learned for imputing missing income log1p transaction amount applied after amount is known Device fingerprint supplied in the scoring API request Expectation that fare amount must be non-negative Vocabulary learned by a one-hot encoder from training categories failed_payments divided by invoices_due with a documented zero-denominator rule Current session country passed from the web request Alert when unknown-category rate doubles week over week
Sign up free — one personalized lesson every day, matched to your role and goals.
Already have an account? Sign in