1.3Standard
Train an image classifier.
Choose models without using the same data for the final performance report.
Choose two correct statements.
× Incorrect
Training data fits parameters, validation data supports model selection, and a held-out test set measures final generalization.
Detailed explanation
CorrectValidation data supports hyperparameter and candidate comparison
It allows development choices without consuming the final test set.
CorrectTraining data fits model parameters
It is used for learning.
IncorrectTrain and report performance on all data
This does not measure performance on unseen data.
IncorrectInspect the test set repeatedly while tuning
This adapts the model to the test set.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01試験ガイドDomain 1.3の訓練・検証・テスト分割を確認する。
Official references
What to confirm
各データ分割の目的とテスト汚染の問題を説明できる。
Key points
- Training
- Validation
- Test
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
Information boundary
Using test results to repeatedly tune a model turns test data into development data.
Unseen evidence
A final estimate is credible only when the model has not adapted to the test examples.
Question IDAWS-AIF-096
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