Oversee AI/ML Model Technique(s) (e.g., Algorithm, Selection) PMI-CPMAI Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   A business needs a model to approve loan applications. Regulators require that the model's decisions be explainable to customers. Which model characteristic is most important to prioritize?

   • AInterpretability over raw performance
   • BMinimal training time
   • CMaximum number of features
   • DHigh complexity for maximum accuracy
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A retail company wants to predict whether a customer will churn in the next 30 days. They have historical data with customer demographics, purchase history, and a binary label indicating churn for past customers. The business team needs a model that is easy to explain to non-technical stakeholders and has moderate accuracy. Which approach should the project manager recommend?

   • AUse a deep neural network to maximize predictive accuracy.
   • BUse reinforcement learning to optimize retention actions.
   • CUse k-means clustering to identify churn segments.
   • DUse logistic regression on the labeled data.
   Reveal Solution Discussion Select an answer first
4. 18application · medium

   A project manager is reviewing the algorithm selection documentation for a predictive maintenance model. The documentation lists the chosen algorithm and its expected accuracy but does not mention why other algorithms were rejected or how the choice aligns with maintenance cost constraints. What should the project manager do?

   • AAsk the team to select a different algorithm immediately.
   • BRequest that the team document the evaluation criteria, alternatives considered, and alignment with cost constraints.
   • CApprove the documentation as is because the algorithm is already chosen.
   • DIgnore the documentation and focus only on the model's accuracy.
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   A company has customer transaction data but no predefined customer segments. They want to discover natural groupings for marketing. Which learning approach is most appropriate?

   • AUnsupervised learning
   • BReinforcement learning
   • CSemi-supervised learning
   • DSupervised learning
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A project manager is reviewing the algorithm selection documentation for a customer segmentation model. The documentation states the chosen algorithm is k-means but does not include the number of clusters or the evaluation metric used. What is the most important gap in the documentation?

   • AThe documentation does not include the cost of the project.
   • BThe documentation lacks the specific hyperparameters and evaluation criteria.
   • CThe documentation does not specify the programming language used.
   • DThe documentation does not list the names of the data scientists.
   Reveal Solution Discussion Select an answer first