Question

A bank uses SageMaker to train a credit approval model. After training, the compliance team wants to verify whether the model approves applications at different rates across gender groups, so they can document and address any unfair outcomes before deployment. Which approach BEST supports this goal?

Accepted Answer

Run SageMaker Clarify to compute post-training bias metrics such as difference in positive proportions in predicted labels across the gender groups

Answer

Enable Amazon Bedrock Guardrails to filter denied topics related to gender from model outputs

Answer

Increase the model's temperature parameter to introduce more variability in approval decisions

Answer

Use SageMaker Clarify SHAP values to explain a single applicant's individual approval decision