1.3Standard
New labeled data arrives daily.
Monitor degradation and retrain when justified.
Choose two operational practices.
× Incorrect
Validate data quality and distribution before retraining, then use fixed tests, production metrics, approval, and staged rollout for updates.
Detailed explanation
CorrectValidate new data quality, distribution, and labels before retraining
Broken data is not used to update the model.
IncorrectReplace production every time data arrives without evaluation
Quality and safety can regress.
CorrectUse fixed evaluation and production metrics and require approval before staged rollout
Retraining regressions can be managed.
IncorrectRecord neither the model nor the data version
Results cannot be reproduced.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 1.3の再学習、モデルレジストリ、評価を確認する。
Official references
What to confirm
定期再学習を自動化しつつ、品質ゲートと承認を残せる。
Key points
- Retraining
- Quality gate
- Approval
Notes
- Environment: AWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
- Compare the definition with the AIF-C01 objective and the service documentation before applying it.
- Keep the concept separate from a particular answer choice or implementation step.
Foundation review
Retraining
Updating a model with new data so it can adapt to changed patterns or requirements.
Quality gate
A required check that data, a model, or a release meets defined quality and safety criteria.
Question IDAWS-AIF-274
EnvironmentAWS公式AIF-C01試験ガイドとAWS公式ドキュメントの確認
Last technical review2026-08-20
Report an error or rights issue