AI Foundations and Technology Ecosystem CRAGE Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   In the AI technology ecosystem, which component is primarily responsible for providing the computational power needed to train large models?

   • AHardware, such as GPUs and TPUs.
   • BAlgorithms, such as neural network architectures.
   • CData, such as training datasets.
   • DPlatforms, such as cloud services.
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A financial institution is deploying an AI system for fraud detection. The system must process transactions in real time and minimize false positives to avoid blocking legitimate customers. The team is considering two deployment options: on-premises servers with high capacity or cloud-based serverless functions. Which factor is most important for meeting the real-time requirement?

   • AData residency requirements
   • BEase of scaling serverless functions
   • CLatency introduced by cloud function cold starts
   • DCost of on-premises hardware
   Reveal Solution Discussion Select an answer first
4. 23expert · hard

   A healthcare AI system is being developed to predict patient readmission risk. The model performs well in testing but fails to generalize when deployed in a different hospital with a different patient population. The team must decide how to proceed. Which action best addresses the root cause?

   • ADeploy the model without changes and monitor performance
   • BRetrain the model using data from the target hospital
   • CUse a different algorithm that is inherently more generalizable
   • DIncrease the model's complexity to capture more patterns
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   A company wants to automate the classification of customer emails into categories like 'complaint', 'inquiry', and 'feedback'. They have a large dataset of labeled emails. Which AI approach is most suitable?

   • ASupervised machine learning, specifically a text classification model.
   • BUnsupervised learning to discover hidden patterns without labels.
   • CReinforcement learning to learn from rewards and penalties.
   • DA rule-based system with manually written rules for each category.
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A company is building a voice assistant for customer service. They need to distinguish between understanding speech (converting audio to text) and understanding intent (determining what the user wants). Which AI subfields are involved?

   • ASpeech recognition (a subfield of NLP) for audio-to-text, and intent recognition (also NLP) for understanding.
   • BComputer vision for speech-to-text and NLP for intent recognition.
   • CRobotics for speech-to-text and deep learning for intent recognition.
   • DNatural language processing (NLP) for both speech-to-text and intent recognition.
   Reveal Solution Discussion Select an answer first