You are using Azure Machine Learning to monitor a trained and deployed model. You implement Event Grid to respond to Azure Machine Learning events. Model performance has degraded due to model input da

The correct answer is C. DatasetDriftDetected. The DatasetDriftDetected event is the correct choice because it is specifically fired when Azure Machine Learning's data drift monitoring service detects that the statistical properties of the model's input data have shifted from the baseline (training data). This is exactly the

Train and deploy models

Question

You are using Azure Machine Learning to monitor a trained and deployed model. You implement Event Grid to respond to Azure Machine Learning events. Model performance has degraded due to model input data changes. You need to trigger a remediation ML pipeline based on an Azure Machine Learning event. Which event should you use?

Options

ARunStatusChanged
BRunCompleted
CDatasetDriftDetected
DModelDeployed

How the community answered

(39 responses)

A
13% (5)
B
5% (2)
C
77% (30)
D
5% (2)

Explanation

The DatasetDriftDetected event is the correct choice because it is specifically fired when Azure Machine Learning's data drift monitoring service detects that the statistical properties of the model's input data have shifted from the baseline (training data). This is exactly the scenario described - model performance degraded due to model input data changes (data drift). RunStatusChanged and RunCompleted relate to the lifecycle of training or pipeline runs, not data quality. ModelDeployed fires when a model is newly deployed, not when performance degrades. Only DatasetDriftDetected is designed to signal that input data distribution has changed, making it the appropriate trigger for a remediation pipeline.

Topics

#Model Monitoring#Data Drift#Azure Machine Learning#Event Grid

Community Discussion

No community discussion yet for this question.

Full DP-100 Practice