Automating Model Retraining PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   In the context of ML deployment, what does continuous delivery (CD) enable?

   • AManual approval for every code change before testing
   • BReal-time monitoring of model predictions
   • CAutomatic retraining of the model on a fixed schedule
   • DAutomated deployment of validated models to staging or production environments
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   What is the role of Cloud Build in an ML CI/CD pipeline?

   • ATo automate the build, test, and deployment steps of the pipeline
   • BTo visualize training metrics
   • CTo monitor model performance in production
   • DTo store the trained model weights
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A company has a model that predicts customer churn. The data changes frequently, and they want to automate the retraining process so that the model is always up-to-date. What is the best way to implement this?

   • AUse a scheduled job that retrains the model every week and deploys it without validation.
   • BUse a manual process where data scientists retrain the model in a notebook and then upload it to a model registry.
   • CUse a CI/CD pipeline that only builds and tests the code, but does not deploy the model.
   • DUse a continuous training pipeline that triggers retraining when new data is available and deploys the model if it passes validation.
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   Which of the following is a benefit of infrequent retraining?

   • ABetter adaptation to rapid shifts in data distribution
   • BHigher model accuracy on recent data
   • CReduced need for monitoring data drift
   • DLower resource consumption and reduced pipeline operational overhead
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A company has a CI/CD pipeline that builds and tests their ML model. They want to automate the deployment of the model to a production endpoint only if it passes all validation checks. What is the best way to achieve this?

   • AAdd a deployment step in the pipeline that runs only if the validation step succeeds, using a conditional in the pipeline configuration.
   • BDeploy the model to production immediately after training, without any validation.
   • CDeploy the model to a staging environment and then manually promote it to production.
   • DUse a manual approval step before deployment, but do not automate it.
   Reveal Solution Discussion Select an answer first