Which of the following best describes how a continuous training pipeline differs from a traditional batch training process?

It requires more human intervention

It only works with deep learning models

It automatically triggers retraining based on new data or drift, rather than waiting for a manual or scheduled run

It uses a larger training dataset

What is the primary goal of a model retraining policy?

To balance model accuracy, operational cost, and business requirements

To ensure the model is retrained only when manually requested

To eliminate the need for model monitoring

To maximize the number of retraining runs regardless of cost

A team is building an automated ML pipeline that includes data validation, model training, and model deployment. They want to ensure that if data validation fails, the pipeline stops and does not proceed to training. What should they use?

A Cloud Function that triggers each step manually.

Vertex AI Pipelines with conditional logic that stops the pipeline if a step fails.

A Kubernetes cron job that runs the steps in parallel.

A single Cloud Build step that runs all the steps in a shell script.

A company has a CI/CD pipeline that deploys a new model to production. They want to monitor the model's performance and automatically rollback if it degrades. However, they are concerned about false positives that could rollback a good model. What is the best way to handle this?

Monitor the model but do not automate rollback; instead, alert the team to decide.

Rollback immediately if any metric drops below a threshold, without any waiting period.

Implement a monitoring system that tracks key metrics and triggers a rollback only if the degradation persists for a certain period or exceeds a threshold.

Use a manual rollback process that requires human approval.

Google CloudProfessional Machine Learning Engineer

Domain 5Objective 2

Automating Model Retraining PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice Questions (Page 2)

Part of the Automating and orchestrating ML pipelines domain, which accounts for ~18% of the PROFESSIONAL-MACHINE-LEARNING-ENGINEER exam.

30questions here

6free pages

9concepts

~18%of the exam

Questions 6–10

6foundation · easy
Which file format is commonly used to define the steps of a Cloud Build pipeline?
Select an answer first
7foundation · easy
Which of the following best describes how a continuous training pipeline differs from a traditional batch training process?
Select an answer first
8foundation · easy
What is the primary goal of a model retraining policy?
Select an answer first
9application · medium
A team is building an automated ML pipeline that includes data validation, model training, and model deployment. They want to ensure that if data validation fails, the pipeline stops and does not proceed to training. What should they use?
Select an answer first
10expert · hard
A company has a CI/CD pipeline that deploys a new model to production. They want to monitor the model's performance and automatically rollback if it degrades. However, they are concerned about false positives that could rollback a good model. What is the best way to handle this?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to PROFESSIONAL-MACHINE-LEARNING-ENGINEER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Google Cloud. “PROFESSIONAL-MACHINE-LEARNING-ENGINEER” is a trademark of its owner, used for identification only.