QuestionQ45
VerifiedML Solution Monitoring, Maintenance, and SecurityAn ML engineer needs to monitor an ML model’s behavior and track bias drift. The engineer must detect issues from an Amazon SageMaker AI endpoint so that actions can be taken to address any identified issues. The engineer has collected Amazon SageMaker Ground Truth labels that show correct results for a set of test data.
Select
Where can the ML engineer view findings and results related to data quality and accuracy drift?
Where can the ML engineer store labels to compare with predictions to measure model quality?
Which service feature can the ML engineer use to monitor bias metrics for a deployed model?
Which service feature can the ML engineer use to roll back to a previous endpoint if a new endpoint's average error rate exceeds a specified threshold?
Community Discussion