Design Use Cases CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A bank has developed a credit risk model to approve loan applications. Before deploying, they want to verify that the model performs consistently across different customer segments and that its decisions align with regulatory requirements. What is the primary use case for testing this model?

   • ATo eliminate the need for ongoing model monitoring
   • BTo increase the speed of model inference
   • CTo ensure the model meets business and compliance requirements
   • DTo reduce the cost of model training
   Reveal Solution Discussion Select an answer first
2. 7expert · hard

   A manufacturing company is testing a predictive maintenance model. They need to ensure it detects rare equipment failures without causing excessive downtime due to false alarms. The cost of a missed failure is high, but false alarms also disrupt production. Which test case design best evaluates this trade-off?

   • AInclude a balanced set of failure and normal operation cases, with emphasis on near-miss conditions
   • BInclude only extreme failure cases
   • CInclude only normal operation cases
   • DInclude only failure cases
   Reveal Solution Discussion Select an answer first
4. 8expert · hard

   A national security agency is testing a model that classifies threat levels from text reports. They need to ensure the model is robust to adversarial text modifications, such as typos or word substitutions, that could evade detection. Which test case design is most appropriate?

   • AInclude only reports with synonyms
   • BInclude only original reports
   • CInclude original reports, reports with typos, and reports with synonyms for key terms
   • DInclude only reports with typos
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   Which of the following best defines a success criterion for model testing?

   • AA measurable threshold that the model must meet to be considered acceptable for deployment
   • BA description of the model's architecture
   • CA general statement that the model should perform well
   • DA list of features that the model should include
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A healthcare startup is developing a model to predict patient readmission risk. They need to define testing objectives that balance clinical accuracy with fairness across different demographic groups. Which set of testing objectives best captures this need?

   • AEnsure the model runs on all devices, supports multiple languages, and has a user-friendly interface
   • BAchieve high accuracy, low latency, and high throughput
   • CMaximize model interpretability, reduce training time, and minimize memory usage
   • DMaximize recall, minimize false positives, and ensure equal performance across groups
   Reveal Solution Discussion Select an answer first