4.1Standard
Deploy a city-trained service nationwide.
Compare user distributions.
Choose two checks.
× Incorrect
Compare training and evaluation data with real users by region, language, device, and conditions, then add missing groups and reevaluate them.
Detailed explanation
CorrectCompare training and evaluation data with real users by region, language, device, and conditions
Distribution bias is visible.
CorrectAdd underrepresented groups and reevaluate quality and safety by group
Representation and impact can improve.
IncorrectMerge small groups into the overall average
Group failures disappear.
IncorrectMore rows automatically guarantee representation
A large dataset can still be biased.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 4.1のデータ代表性、公平性、評価を確認する。
Official references
What to confirm
利用者分布から外れた評価データを見つけ、集団別に再評価できる。
Key points
- Representation
- Distribution
- Group evaluation
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
Data representativeness
How well data reflects the people, conditions, and cases a system is meant to serve.
Group evaluation
Measuring quality, safety, or outcomes separately for defined populations or conditions.
Question IDAWS-AIF-361
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