Artificial Intelligence Fundamentals GOAA Practice Questions (Page 10)

Questions 46–48

1. 46application · medium

   A security operations center (SOC) wants to build an AI system that flags phishing emails. They have a large dataset of emails labeled 'phishing' or 'benign'. The team needs to train a model that learns the mapping from email content to labels. Which approach should they use?

   • ADeep learning only
   • BSupervised learning
   • CUnsupervised learning
   • DReinforcement learning
   Reveal Solution Discussion Select an answer first
2. 47expert · hard

   A team is building an AI system that plays chess. They have a large database of historical chess games with outcomes. The system should learn to predict the best move from a given board position. Which approach is most appropriate?

   • AReinforcement learning with self-play
   • BUnsupervised learning to cluster board positions
   • CSupervised learning on historical games to predict moves
   • DDeep learning with a generative model
   Reveal Solution Discussion Select an answer first
3. 48application · medium

   A startup is developing an AI that generates realistic fake audio of a CEO's voice for a social engineering simulation. They use a deep learning model that learns to produce new audio samples from a training set of the CEO's voice recordings. Which type of neural network architecture is most commonly associated with this generative task?

   • AConvolutional neural network (CNN) for classification
   • BFeedforward neural network for regression
   • CRecurrent neural network (RNN) for sentiment analysis
   • DGenerative adversarial network (GAN)
   Reveal Solution Discussion Select an answer first