Explain Core AI Principles and Terminology CY0-001 Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A security vendor describes their new product as 'AI-powered' because it uses a simple rule-based system to block known malicious IP addresses. Which statement best characterizes this product?

   • AIt is not true AI because it does not learn or adapt; it only follows predefined rules
   • BIt is a strong example of deep learning because it processes IP addresses automatically
   • CIt is a form of unsupervised learning because it categorizes IPs as malicious or not
   • DIt is an example of reinforcement learning because it takes action to block traffic
   Reveal Solution Discussion Select an answer first
2. 7expert · hard

   A security team is developing an AI model to prioritize security alerts. They train the model on historical alert data, but they discover that the model gives lower priority to alerts from certain departments, even when the alerts are equally severe. This is because those departments had fewer historical incidents. What is the most appropriate first step to address this issue?

   • ARe-balance the training data or adjust the model to reduce bias, and monitor for fairness
   • BRemove all alerts from those departments from the training data
   • CIgnore the issue because the model is only for prioritization, not for access control
   • DIncrease the model's complexity to better fit the data
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A security analyst wants to use machine learning to identify new, unknown malware families. They have a large collection of malware samples but no labels indicating which family each sample belongs to. Which machine learning approach is most suitable for this task?

   • ASupervised learning with a classifier trained on known families
   • BUnsupervised learning using clustering to group similar malware samples
   • CSemi-supervised learning using only the labeled portion
   • DReinforcement learning with a reward based on detection accuracy
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   What is AI bias in the context of cybersecurity models?

   • AThe tendency of a model to overfit to the training data.
   • BA type of encryption used to protect the model.
   • CA measure of how fast the model processes data.
   • DA systematic error in the model's predictions that results from biased training data or flawed algorithms.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A security team is comparing two models for detecting malware: a traditional machine learning model using hand-crafted features (e.g., file size, entropy) and a deep learning model that takes raw bytes as input. What is a key advantage of the deep learning model in this context?

   • AIt requires less training data than traditional machine learning
   • BIt is always more interpretable than traditional machine learning
   • CIt does not require any computational resources for training
   • DIt can automatically learn relevant features from the raw data, reducing the need for manual feature engineering
   Reveal Solution Discussion Select an answer first