A company has used Amazon SageMaker to deploy a predictive ML model in production. The company is using SageMaker Model Monitor on the model. After a model update, an ML engineer notices data…

Question

Options

• AAdjust the model's parameters and hyperparameters.
• BInitiate a manual Model Monitor job that uses the most recent production data.
• CCreate a new baseline from the latest dataset. Update Model Monitor to use the new baseline for
• DInclude additional data in the existing training set for the model. Retrain and redeploy the model.

• A
  16% (6)
• B
  8% (3)
• C
  70% (26)
• D
  5% (2)

Explanation

After a model update, the existing baseline used by Model Monitor reflects the old model's expected data distribution - so violations are expected and the baseline itself is now stale. Option C is correct because creating a new baseline from the latest dataset recalibrates Model Monitor's reference point to match the updated model's expected inputs and outputs, eliminating false-positive quality alerts.

• A is wrong because adjusting parameters/hyperparameters addresses model accuracy, not the mismatch between the monitoring baseline and new data distribution.
• B is wrong because running a manual job still uses the outdated baseline - it re-checks data against the wrong reference and doesn't fix the root cause.
• D is wrong because retraining addresses model performance degradation, not a baseline alignment issue; the data quality alerts stem from the stale baseline, not bad training data.

Memory tip: Think of the baseline as a "contract" the model signs with Monitor - when you update the model, you must renew the contract (new baseline) so Monitor knows what "normal" looks like for the new version.

No community discussion yet for this question.