Determine If Data Meets Solution Needs PMI-CPMAI Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A financial services company is building a model to detect money laundering. The data science team has transaction data, but they lack data on customer risk profiles, which the business considers essential. The team is identifying data gaps. What should they do?

   • AProceed with transaction data only and rely on anomaly detection techniques.
   • BUse transaction data to infer customer risk profiles through modeling.
   • CIdentify the risk profile data gap and explore internal sources, such as KYC records, or external data providers.
   • DAsk the business to remove the risk profile requirement from the model's scope.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A ride-sharing company is building a model to predict ride demand. The training data is from a major city, but the model will be deployed in smaller cities. The data science team is validating representativeness. What should they do?

   • AProceed with the data because the model can learn general demand patterns.
   • BValidate whether the training data is representative of the smaller cities' demand patterns.
   • CUse the major city data and add a city size feature to the model.
   • DCollect data from the smaller cities before training.
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   What does a data sufficiency assessment primarily evaluate?

   • AWhether the data volume and quality are enough to train a robust AI model
   • BWhether the data is labeled correctly for supervised learning
   • CWhether the data is stored in a cloud or on-premises environment
   • DWhether the data contains personally identifiable information
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   A data scientist is evaluating a dataset with only 100 records for a deep learning model that typically requires millions. Which aspect of data sufficiency is most directly challenged?

   • AData veracity
   • BData variety
   • CData velocity
   • DData volume
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   An e-commerce company is building a fraud detection model. They have a large dataset of transactions, but only 0.1% are fraudulent. The model will be used to flag transactions in real time. The team is considering whether the data is sufficient. The business has a low tolerance for false positives because they disrupt legitimate customers. What is the most important consideration for the data sufficiency assessment?

   • AThe data is sufficient because the model can be trained to minimize false positives.
   • BThe data is sufficient because it includes both fraudulent and legitimate transactions.
   • CThe total number of transactions is large enough to train a deep learning model.
   • DThe number of fraudulent examples is sufficient to train a model that can detect fraud with high recall.
   Reveal Solution Discussion Select an answer first