Artificial Intelligence Fundamentals GOAA Practice Questions (Page 3)

Questions 11–15

1. 11application · easy

   A bank wants to predict whether a loan applicant will default. They have historical data with features like income, credit score, and loan amount, and each record is labeled 'default' or 'no default'. Which machine learning task is this?

   • ARegression
   • BUnsupervised clustering
   • CReinforcement learning
   • DSupervised classification
   Reveal Solution Discussion Select an answer first
2. 12expert · hard

   A city is implementing an AI system to optimize traffic light timings. The system must adapt to real-time traffic conditions and learn from the outcomes of its decisions. The goal is to minimize average commute time. Which AI approach is most suitable?

   • AUnsupervised learning to cluster traffic patterns
   • BReinforcement learning
   • CSupervised learning on historical traffic data
   • DA rule-based expert system
   Reveal Solution Discussion Select an answer first
4. 13expert · hard

   A security team is deploying an AI-based intrusion detection system (IDS) that uses a deep learning model trained on network traffic data. The model achieves high accuracy on the test set but fails to detect novel attacks that are different from the training data. The team needs to improve the system's ability to handle unseen attack patterns. Which approach is most effective?

   • AIncorporate anomaly detection techniques that flag deviations from normal behavior
   • BSwitch to a rule-based signature system
   • CIncrease the model's complexity by adding more layers
   • DUse a larger dataset of known attacks only
   Reveal Solution Discussion Select an answer first
5. 14expert · medium

   A company wants to deploy an AI system that can recognize faces in real-time video streams for physical security. They have a limited amount of labeled face data. Which approach is most likely to achieve good performance with limited data?

   • AUse unsupervised clustering on the video frames
   • BTrain a deep neural network from scratch on the limited data
   • CUse a pre-trained deep learning model and fine-tune it on the company's data
   • DUse a simple rule-based system
   Reveal Solution Discussion Select an answer first
6. 15application · medium

   A data science team is building a system to detect malware from raw executable bytes. They have millions of unlabeled binary files. They want the model to automatically learn hierarchical features, from low-level byte patterns to high-level behavioral abstractions, without manual feature engineering. Which approach is most appropriate?

   • AReinforcement learning with a policy network
   • BClassical machine learning with handcrafted features
   • CDeep learning with a neural network
   • DUnsupervised clustering on raw bytes
   Reveal Solution Discussion Select an answer first