A fraud detector should reduce missed fraud.
A home-price model should reduce the difference between predictions and actual values.
Choose two correct task-metric pairs.
Use recall to assess missed fraud and a regression error such as mean absolute error for price predictions.
Detailed explanation
Class imbalance can hide missed fraud in overall accuracy.
A confusion matrix is a classification tool, not a direct continuous-value metric.
Mean absolute error measures the average absolute difference from the actual value.
Recall measures how much actual fraud is detected.
Check official documentation
Compare the following points with the official documentation.
AWS公式AIF-C01ガイドDomain 1.3のモデル評価指標を確認する。
Official references
What to confirm
分類と回帰で指標を分け、業務上重要な誤りを指標へ反映できる。
Key points
- Metric selection
- Missed positives
- Regression error
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
Error cost
Missing a real case and raising a false alarm can have different consequences.
Metric fit
A metric must reflect the output type and the decision risk rather than be chosen by habit.