1.2Standard
Synthetic examples are added for a rare class.
Manage training risk.
Choose two checks.
× Incorrect
Compare distributions, boundaries, duplicates, leakage, protected attributes, and privacy with real data, and evaluate on a real validation set.
Detailed explanation
CorrectEvaluate performance, fairness, and privacy on a real-data validation set
Real-world generalization is measured.
IncorrectDuplicate synthetic rows in train and test without limits
Evaluation becomes optimistic.
CorrectCompare distribution, boundary cases, duplicates, leakage, and protected attributes with real data
Quality and bias are tested.
IncorrectArtificial data is always anonymous and needs no review
Reidentification and bias can remain.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 1.2の合成データ、品質、公平性、プライバシーを確認する。
Official references
What to confirm
合成データを無条件に信頼せず、現実データで有効性とリスクを確認できる。
Key points
- Synthetic data
- Reidentification
- Validation
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
Synthetic data
Artificially generated records designed to resemble characteristics of real data.
Re-identification risk
The possibility that a record can be linked back to a person or source.
Question IDAWS-AIF-453
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