Objective 1.5 Communicate with Stakeholders AIP-210 Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   A data scientist needs to explain the concept of 'overfitting' to a non-technical project manager. Which explanation is most accessible?

   • AThe model memorizes the training data too well and fails to generalize to new data
   • BThe model's hyperparameters are not properly tuned, causing instability
   • CThe model has high variance and low bias, leading to poor out-of-sample performance
   • DThe loss function is minimized on the training set but not on the validation set
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A data science team is explaining a new recommendation engine to the marketing team. The marketing team is confused about why the engine sometimes recommends products the customer has never viewed. Which explanation is most accessible and accurate?

   • AThe engine is broken and needs to be fixed.
   • BThe engine is random, so sometimes it suggests products by chance.
   • CThe engine uses a deep neural network with 12 layers and a softmax output layer.
   • DThe engine finds patterns among similar customers, so it can suggest items that other people with similar tastes liked.
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A company is deploying an AI chatbot for customer support. The support team expects the chatbot to handle all inquiries without human intervention. The project manager knows the chatbot will only handle common questions and will escalate complex issues. What is the best way to manage the support team's expectations?

   • APromise that the chatbot will be upgraded to handle all inquiries within a month.
   • BLet the support team discover the limitations after the chatbot is deployed.
   • CClearly communicate the chatbot's capabilities and limitations, and provide a roadmap for future improvements.
   • DTell the support team that the chatbot is only a temporary tool and will be removed soon.
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A bank is developing an AI system to approve loan applications. The compliance team is concerned about regulatory requirements, while the sales team wants faster approvals. The data science team is focused on model accuracy. What is the best way to address these differing perspectives?

   • ATell the compliance team that their concerns are unfounded because the model is fair.
   • BLet the data science team decide the final approach because they understand the model best.
   • CHold a joint workshop where each team presents its priorities and constraints, then work together to find a balanced solution.
   • DPrioritize the sales team's need for speed because they generate revenue.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A cross-functional team is building an AI system to recommend personalized learning paths for students. The team includes data scientists, educators, and IT staff. During development, the educators express concerns that the recommendations may not align with the curriculum. What is the best way to facilitate collaboration?

   • ALimit communication to written reports to avoid conflicts.
   • BAsk the educators to wait until the system is fully built before giving feedback.
   • CTell the educators that the data scientists will handle all decisions about the recommendation logic.
   • DSchedule regular feedback sessions where educators can review model outputs and provide input on the recommendation logic.
   Reveal Solution Discussion Select an answer first