Automating Model Retraining PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A team is new to CI/CD and wants to apply it to their ML projects. They have a repository with training code and a model evaluation script. What is the first thing they should automate?

   • ADeploying the model to production automatically after every commit.
   • BRunning the evaluation script as a test in the CI pipeline to ensure the code changes don't break the model's performance.
   • CCreating a Cloud Build trigger that deploys the model to a staging environment.
   • DSetting up a continuous training pipeline that retrains the model on a schedule.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   What is the primary benefit of automating model deployment with CI/CD tools?

   • AIt reduces the risk of human error and ensures consistent deployment steps
   • BIt eliminates the need for model validation
   • CIt ensures the model is trained on the latest data automatically
   • DIt guarantees the model will have zero latency in production
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which Google Cloud service is designed for orchestrating ML pipeline steps?

   • AVertex AI Pipelines
   • BCloud Functions
   • CCloud Pub/Sub
   • DCloud Storage
   Reveal Solution Discussion Select an answer first
5. 14application · medium

   An e-commerce company has a recommendation model that needs to be updated frequently as user behavior changes. They want to automate the entire process from data ingestion to model deployment. What should they implement?

   • AA scheduled batch job that retrains the model every night and deploys it without validation.
   • BA manual process where data scientists retrain the model in a notebook and then upload it to a model registry.
   • CA continuous training (CT) pipeline that automatically retrains the model when new data is available and deploys it if it passes validation.
   • DA CI/CD pipeline that only builds and tests the code, but does not deploy the model.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   What is the main goal of a continuous training (CT) pipeline?

   • ATo deploy the model without any testing
   • BTo train the model only on the initial dataset
   • CTo automatically retrain and redeploy models when new data becomes available
   • DTo manually retrain the model once a year
   Reveal Solution Discussion Select an answer first