Specialized Testing Considerations CT-AI Practice Questions (Page 6)

Questions 26–30

1. 26foundation · easy

   What is a common strategy for handling concept drift in an AI-based system?

   • ARetraining the model with recent data that reflects the current concept.
   • BDeleting all historical data to avoid confusion.
   • CIgnoring the drift and hoping the model recovers.
   • DIncreasing the model's complexity without retraining.
   Reveal Solution Discussion Select an answer first
2. 27application · medium

   A team is monitoring an AI system that detects fraudulent transactions. They want to detect concept drift early to avoid significant financial losses. Which metric would be most indicative of concept drift?

   • AA change in the average transaction amount.
   • BA decrease in the model's training time.
   • CAn increase in the number of transactions processed by the system.
   • DA significant drop in the model's precision and recall on recent transactions compared to historical performance.
   Reveal Solution Discussion Select an answer first
4. 28foundation · easy

   Which of the following is a key piece of information that should be documented for an AI component?

   • AThe number of meetings held during development.
   • BThe brand of the hardware used by the development team.
   • CThe personal preferences of the development team.
   • DThe training data sources and preprocessing steps.
   Reveal Solution Discussion Select an answer first
5. 29foundation · easy

   What is the primary purpose of using a standard documentation format or template for AI components?

   • ATo make the documentation longer and more detailed.
   • BTo replace the need for testing the AI component.
   • CTo hide sensitive information about the AI model.
   • DTo ensure consistency and completeness of documentation across different AI components.
   Reveal Solution Discussion Select an answer first
6. 30foundation · easy

   In which scenario is automation bias most likely to lead to an incorrect decision in an AI-based system?

   • AAn AI system is tested with a diverse set of test cases.
   • BA doctor accepts an AI diagnostic suggestion without checking the patient's conflicting symptoms.
   • CA tester documents all AI outputs for later review.
   • DA user manually reviews all AI recommendations before acting on them.
   Reveal Solution Discussion Select an answer first