1.2Standard
A churn model includes a field recorded after churn.
Keep only data available at prediction time.
Choose two responses.
× Incorrect
Remove future information by auditing feature timing and reassess with time-based splits.
Detailed explanation
IncorrectKeep any highly correlated field even if it arrives later
It uses future information.
CorrectAudit source timestamps and update rules and reassess with time-based splits
Inflated estimates can be detected.
CorrectKeep only features available by the prediction timestamp
Future information is excluded.
IncorrectIgnore feature availability when test accuracy is high
Overestimation remains.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 1.2のデータ漏洩、特徴量生成時点、評価を確認する。
Official references
What to confirm
特徴量の相関ではなく、予測時点で利用可能かを確認してモデルを評価できる。
Key points
- Target leakage
- Prediction time
- Time split
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
Target leakage
A feature contains information about the target that would not be known when predicting.
Point-in-time correctness
Ensuring each feature value reflects only information available at the prediction timestamp.
Question IDAWS-AIF-415
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