Transparency, Interpretability and Explainability CT-AI Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   Which of the following is a common technique used to achieve interpretability in AI models?

   • AUsing a random forest with hundreds of trees
   • BUsing a support vector machine with a radial basis function kernel
   • CUsing a deep neural network with many hidden layers
   • DUsing a decision tree model
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   What is an ethical consideration related to the lack of explainability in AI systems?

   • AIt can lead to unfair or discriminatory outcomes that are difficult to detect and challenge
   • BIt increases the cost of the system, which is unethical
   • CIt always results in lower accuracy, which is unethical
   • DIt makes the system more secure, which is unethical
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   In which scenario is 'local interpretability' most useful?

   • AWhen a regulator wants to audit the model's compliance with data protection laws
   • BWhen a data scientist wants to understand which features are most important across the entire model
   • CWhen a developer wants to compare the overall performance of two different models
   • DWhen a loan applicant wants to know why their specific application was rejected
   Reveal Solution Discussion Select an answer first
5. 24expert · hard

   A multinational corporation is deploying an AI system for employee performance evaluation. The system will be used in the EU, where the GDPR grants individuals the right to not be subject to decisions based solely on automated processing without meaningful human involvement. The company's HR department wants to fully automate the process to save time. The legal department insists on compliance. What is the most appropriate way to proceed?

   • AFully automate the process but only for employees who sign a waiver agreeing to automated decisions.
   • BFully automate the process and provide employees with a general statement that the system is fair and unbiased.
   • CImplement a process where the AI system provides a recommendation and a human manager makes the final decision, with the ability to review the AI's reasoning.
   • DFully automate the process and provide each employee with a detailed technical explanation of how the AI model works.
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   What is the primary role of 'explainability' in an AI-based system?

   • ATo ensure the system's predictions are always correct
   • BTo make the system's source code available for public review
   • CTo automatically fix errors in the model's predictions
   • DTo provide reasons or justifications for the system's decisions in a way that humans can understand
   Reveal Solution Discussion Select an answer first