AI Project Life Cycle, MLOps, and DataOps CAIPM Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   In MLOps, what is the primary purpose of continuous integration (CI) for machine learning projects?

   • ATo store and version training datasets
   • BTo automatically test and validate code and model components whenever changes are made
   • CTo deploy models to production on a fixed schedule
   • DTo monitor model performance in real time
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A company wants to implement MLOps for its computer vision model. The team needs to automate the training, validation, and deployment of the model. They also need to track the performance of the model in production. Which set of tools would best support these requirements?

   • AA CI/CD tool (e.g., Jenkins) and a model monitoring tool (e.g., Prometheus)
   • BA model registry (e.g., MLflow) and a data catalog (e.g., Amundsen)
   • CA containerization tool (e.g., Docker) and a data visualization tool (e.g., Tableau)
   • DA data orchestration tool (e.g., Airflow) and a version control system (e.g., Git)
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A telecom company has a churn-prediction model that is retrained monthly. The ML team wants to implement continuous delivery so that new model versions can be deployed quickly, but they are worried about downtime during deployment. Which deployment strategy would allow the team to deploy a new model version without interrupting service?

   • ARolling deployment
   • BBlue-green deployment
   • CRecreate deployment
   • DCanary deployment
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which type of tool is specifically used in DataOps to schedule and coordinate data processing tasks across multiple systems?

   • AData orchestration platforms
   • BModel registries
   • CVersion control systems
   • DCI/CD pipelines
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A healthcare company is developing an AI diagnostic tool. The model shows high accuracy in the lab, but the deployment team is concerned about data privacy and regulatory compliance. The project is in the deployment phase. Which action is MOST critical to ensure a successful and compliant deployment?

   • AConduct a thorough validation of the model on real-world data and ensure compliance with data protection regulations.
   • BDeploy the model quickly to gain a competitive advantage.
   • CFocus only on model performance metrics and ignore data privacy.
   • DAsk the data science team to improve the model's accuracy further.
   Reveal Solution Discussion Select an answer first