1.4 AI/Machine Learning 400-007 Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A healthcare provider is considering an AI system to automate the triage of patient messages. The system would reduce the workload of nurses, but it also introduces risks related to misdiagnosis and patient safety. The provider wants to ensure that the AI system is used responsibly and that patients are not harmed. Which governance framework should the provider implement?

   • AUse a third-party AI vendor and transfer all liability to the vendor.
   • BEstablish a multidisciplinary oversight committee that includes clinicians, ethicists, and patient representatives to review the AI system's performance and policies.
   • CLimit the AI system to non-critical tasks and avoid any human oversight to reduce costs.
   • DDeploy the AI system and rely on post-deployment monitoring to catch any issues.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A retail chain wants to use AI to optimize inventory management across its stores. The company has limited historical data for some new product lines. The CEO asks the design team to justify the AI investment. Which approach best articulates the business need and expected ROI?

   • AFocus on the potential for AI to reduce stockouts and overstock, and estimate savings based on industry benchmarks.
   • BHighlight the technical sophistication of the AI model as the primary justification.
   • CPromise that AI will eliminate all inventory costs within the first year.
   • DDefer the AI project until more historical data is collected, to ensure accurate predictions.
   Reveal Solution Discussion Select an answer first
4. 13application · medium

   A global logistics company uses an AI model to optimize delivery routes. The model is trained on data from multiple countries, and the company must comply with different data protection regulations. The company wants to ensure that the model does not inadvertently leak sensitive information from one region to another. What should the company do?

   • AAnonymize all data before training to eliminate any privacy concerns.
   • BStore all training data in a central location and apply access controls.
   • CTrain separate models for each region using only that region's data.
   • DUse a single global model and rely on encryption to protect data in transit.
   Reveal Solution Discussion Select an answer first
5. 14expert · hard

   A defense contractor is developing an AI model for object detection in satellite imagery. The model is trained on classified data and must be protected from adversarial attacks that could cause misclassification. The contractor also needs to ensure that the training data has not been poisoned. Which security measures should the contractor implement?

   • ARely on the classified network's physical security to protect the model and data.
   • BEncrypt the model and training data at rest and in transit.
   • CUse a simpler model that is less susceptible to adversarial attacks.
   • DUse adversarial training to make the model robust to attacks, and implement data provenance checks to detect poisoning.
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   What is the primary goal of auto-scaling for AI/ML infrastructure?

   • ATo adjust resources dynamically to match workload demand while maintaining performance
   • BTo reduce the complexity of the AI/ML pipeline
   • CTo ensure that all resources are always fully utilized
   • DTo eliminate the need for capacity planning
   Reveal Solution Discussion Select an answer first