Feature Stores Quiz

Answers and explanations appear as you take the quiz.

1. Why use shared feature definitions for training and online serving?
   • To guarantee every cached value is always fresh
   • To reduce skew caused by separate transformation implementations
   • To eliminate the need for timestamps
   • To choose a GPU model automatically
2. What is the difference between the offline and online stores?
   • Offline holds historical data for training; online serves the latest values fast for live predictions
   • Offline is for images, online is for text
   • They are identical with different names
   • Offline runs on weekends, online runs on weekdays
3. Why are point-in-time joins important when building training data?
   • They make the database run faster
   • They translate features into other languages
   • They prevent data leakage by using feature values as they were at each event's time
   • They reduce the model's file size
4. Which storage pattern commonly supports low-latency feature lookup by entity key?
   • An archival tape read for every request
   • A printed document catalog
   • A model-weight-only file
   • A key-value online store such as Redis or DynamoDB
5. What is a 'feature' in this context?
   • An input variable derived from data, like average spend over 30 days
   • A bug in the model code
   • The color scheme of a dashboard
   • A type of GPU
6. What extra benefit does a shared feature catalog provide?
   • It deletes old data automatically every hour
   • It replaces the need for any model training
   • It lets features built for one model be discovered and reused by other models
   • It encrypts the company's email
7. A feature’s event time precedes a loan decision, but the value only arrived afterward. What must a historical training join consider?
   • Only whether its column name is short
   • A rule that every past event was already known
   • Only the current online value
   • Whether that value was actually available at prediction time
8. An online feature has not been refreshed after new events arrived. Which property should a feature-store monitor inspect?
   • The font size of the registry page
   • Feature freshness relative to the prediction’s needs
   • The model’s number of attention heads only
   • Whether the feature name contains a digit