Identify the Differences and Advantages of Model Serving Approaches: Batch, Realtime, and Streaming MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

1. 6foundation · easy

   Which type of application benefits most from realtime model serving?

   • AA dashboard that updates every hour with new predictions
   • BA nightly job that generates a report for the marketing team
   • CA system that scores sensor data as it arrives to detect anomalies
   • DAn interactive chatbot that needs to respond to user messages instantly
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   What is a key advantage of streaming model serving?

   • AIt is the most cost-effective for processing a large, finite dataset
   • BIt provides the lowest possible latency for a single request
   • CIt enables continuous processing of unbounded data streams with near-real-time results
   • DIt is simpler to implement than batch serving
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   Which statement best describes batch model serving?

   • AProcessing individual requests as they arrive to provide immediate predictions
   • BRunning predictions on a large collection of records in a scheduled, offline job
   • CContinuously scoring data records as they stream into a system
   • DDeploying a model to a REST endpoint for on-demand calls from an application
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   How does batch serving typically compare to realtime serving in terms of latency and cost?

   • ABatch is always cheaper and faster than realtime
   • BBatch and realtime have similar latency and cost
   • CBatch has lower latency but is more expensive
   • DBatch has higher latency but is often more cost-efficient for large volumes
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A social media platform wants to flag hate speech in user posts. Posts arrive continuously 24/7, and moderation should happen within a few seconds of posting so harmful content is removed quickly. The volume is unbounded and varies with user activity. Which serving approach is most appropriate?

   • ARealtime serving with a REST endpoint called for each post
   • BBatch serving that runs hourly on all posts from the last hour
   • CBatch serving that runs every minute on new posts
   • DStreaming serving that processes posts as they arrive in a message queue
   Reveal Solution Discussion Select an answer first