Google Cloud Professional Machine Learning Engineer — Question 189

Your team has a model deployed to a Vertex AI endpoint. You have created a Vertex AI pipeline that automates the model training process and is triggered by a Cloud Function. You need to prioritize keeping the model up-to-date, but also minimize retraining costs. How should you configure retraining?

Answer options

• A. Configure Pub/Sub to call the Cloud Function when a sufficient amount of new data becomes available
• B. Configure a Cloud Scheduler job that calls the Cloud Function at a predetermined frequency that fits your team’s budget
• C. Enable model monitoring on the Vertex AI endpoint. Configure Pub/Sub to call the Cloud Function when anomalies are detected
• D. Enable model monitoring on the Vertex AI endpoint. Configure Pub/Sub to call the Cloud Function when feature drift is detected

Correct answer: D

Explanation

The correct answer is D because detecting feature drift indicates that the model's input data characteristics have changed, which may necessitate retraining to maintain accuracy. While option A focuses on retraining based on new data volume, it may not reflect the need for model updates. Option B follows a fixed schedule that might not be efficient in terms of model performance and costs. Option C addresses anomalies but does not specifically target the underlying changes in the data distribution that feature drift represents.