Objective 7.1 Report Findings CERTIFIED-DATA-SCIENCE-PRACTITIONER Practice Questions (Page 12)

Questions 56–60

1. 56application · medium

   A data scientist has a trained sentiment analysis model and needs to show product managers a working demo where they can paste a product review and see the predicted sentiment. The demo must be a simple web page, and the data scientist has limited web development experience. Which implementation is the most straightforward?

   • ABuild a Django app with a full database-backed model to store each review and prediction.
   • BExpose the model as a REST API and ask product managers to test it using curl commands.
   • CCreate a static HTML page that uses a JavaScript library to load the model and run inference in the browser.
   • DCreate a Flask app with a route that renders an HTML form, accepts the review text via POST, calls the model, and displays the result.
   Reveal Solution Discussion Select an answer first
2. 57foundation · easy

   Which of the following is typically included in model lifecycle documentation?

   • AThe model version, training data sources, and evaluation metrics
   • BThe list of all employees in the organization
   • CThe marketing budget for the product using the model
   • DThe personal login credentials of the data science team
   Reveal Solution Discussion Select an answer first
4. 58foundation · easy

   When explainability is critical, why might a data scientist prefer an interpretable model over a black-box model?

   • AInterpretable models require less data
   • BInterpretable models always have higher accuracy
   • CInterpretable models are always faster to train
   • DInterpretable models are easier to explain to stakeholders and regulators
   Reveal Solution Discussion Select an answer first
5. 59foundation · easy

   How can attention weights be used to explain a model's prediction?

   • ABy showing the model's architecture
   • BBy showing which input tokens were given the most weight
   • CBy showing the model's confidence in its prediction
   • DBy showing the training loss over time
   Reveal Solution Discussion Select an answer first
6. 60foundation · easy

   What is a key advantage of using a decision tree for explainability?

   • AIt always achieves higher accuracy than other models
   • BThe decision path for each prediction can be easily traced and understood
   • CIt can handle only numerical features
   • DIt requires no feature scaling
   Reveal Solution Discussion Select an answer first