Types of AI - ML, DL, Generative AI, and Agents CAIPM Practice Questions (Page 8)

Questions 36–37

1. 36application · medium

   A healthcare provider wants to generate synthetic patient records to share with researchers without exposing real patient data. The synthetic records must preserve statistical patterns of the original data. Which approach is most appropriate?

   • AApply a deep learning classifier to categorize records by diagnosis
   • BUse a supervised ML model to predict patient outcomes from the records
   • CDeploy an AI agent to automatically de-identify and share real records
   • DTrain a generative adversarial network (GAN) on the original records
   Reveal Solution Discussion Select an answer first
2. 37expert · hard

   A logistics company wants to optimize delivery routes in real time. They have historical delivery data and live traffic feeds. The system must adapt to unexpected road closures and also provide explanations for route changes to dispatchers. Which approach is most appropriate?

   • AUse a generative AI model to create new routes from scratch each time.
   • BUse a deep learning model alone to predict routes without any decision-making component.
   • CUse a supervised ML model to predict delivery times and an AI agent to adjust routes dynamically based on live conditions.
   • DUse a rule-based system with fixed routing rules and no learning.
   Reveal Solution Discussion Select an answer first