Explain Core AI Principles and Terminology CY0-001 Practice Questions (Page 1)

Questions 1–5

1. 1foundation · easy

   What is the main difference between supervised and unsupervised learning?

   • ASupervised learning requires a neural network, while unsupervised learning does not.
   • BSupervised learning uses labeled data, while unsupervised learning finds patterns in unlabeled data.
   • CSupervised learning is faster than unsupervised learning.
   • DSupervised learning is used for clustering, while unsupervised learning is used for classification.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A security operations center (SOC) wants to develop an AI agent that automatically adjusts firewall rules in response to an ongoing attack, learning from the outcomes of its actions (e.g., whether the attack was blocked or the network remained available). Which machine learning paradigm is most appropriate for this dynamic, action-oriented environment?

   • ADeep learning with a convolutional neural network
   • BUnsupervised learning
   • CReinforcement learning
   • DSupervised learning
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A cybersecurity firm is building a model to detect zero-day malware by analyzing raw executable file bytes. They have a very large dataset of benign and malicious binaries. Which approach is most likely to automatically learn hierarchical features (e.g., from byte patterns to higher-level structures) without extensive manual feature engineering?

   • AA simple linear regression model
   • BDeep learning with a neural network with multiple hidden layers
   • CTraditional machine learning with hand-crafted features
   • DA decision tree with a single split
   Reveal Solution Discussion Select an answer first
5. 4expert · hard

   A security team is deploying a neural network to detect malware. They are concerned about adversarial attacks where attackers modify malware to evade detection. What is a key limitation of neural networks in this context?

   • ANeural networks can be fooled by small, carefully crafted perturbations to the input that are imperceptible to humans
   • BNeural networks are immune to adversarial attacks because they learn complex patterns
   • CNeural networks cannot process binary data such as executable files
   • DNeural networks require too much memory to store adversarial examples
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A security team wants to develop an AI system that automatically responds to intrusions by taking actions such as blocking IPs or isolating hosts. The system should learn from the outcomes of its actions (e.g., whether the threat was neutralized or the network was disrupted). Which approach is most suitable, and what is a key challenge?

   • AUnsupervised learning, but it requires manual labeling of all actions
   • BReinforcement learning, but it requires careful design of the reward function to avoid unintended consequences
   • CDeep learning, but it requires a GPU cluster for training
   • DSupervised learning, but it requires a large labeled dataset of actions and outcomes
   Reveal Solution Discussion Select an answer first