Resampling reduced a recall gap between groups.
Check side effects before and after deployment.
Choose two appropriate validations.
Compare multiple group and overall metrics before and after the intervention and consider legal requirements and appeal outcomes.
Detailed explanation
One improved metric can hide a new degradation.
One improved metric can hide a new degradation.
Numbers alone may not capture real-world impact.
Numbers alone may not capture real-world impact.
Other harms may be hidden.
Other harms may be hidden.
Data and usage can change fairness over time.
Data and usage can change fairness over time.
Try it yourself
An example you can run in a temporary verification environment.
AWS公式AIF-C01 Domain 4.1の公平性評価と責任あるAI運用を確認する。Expected result
改善指標だけでなく副作用、法令、異議申立てを含めて検証できる。Key points
- Before/after
- Side effects
- Continuous review
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
- Command output formatting can vary slightly by distribution or tool version.
- Run the example in a temporary directory or process when possible.
Foundation review
Read the scope first
Check whether the command acts on the current shell, a new process, an existing process, or a file.
Verify the observable result
Use the supplied command and compare the output with the expected result.