1.2Standard
Fraud is only 0.1 percent of transactions.
Measure misses and false alarms.
Choose two evaluations.
× Incorrect
For imbalanced data, use precision, recall, F1, PR curves, and a confusion matrix together with business costs.
Detailed explanation
IncorrectRemove the minority class from testing
Important errors are hidden.
IncorrectTreat high accuracy as proof that all fraud is found
The majority class inflates it.
CorrectUse a confusion matrix and PR curve at the operating threshold
The decision is concrete.
CorrectCheck precision, recall, F1, and the business cost of misses and false alarms
Minority-class performance is visible.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01 Domain 1.2の不均衡データ、混同行列、適合率、再現率を確認する。
Official references
What to confirm
少数クラスの検出性能を業務コストに沿って評価できる。
Key points
- Imbalance
- Recall
- Precision
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
Precision
The share of predicted positives that are actually positive.
Recall
The share of actual positives that the model identifies.
Question IDAWS-AIF-432
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