What is the purpose of configuring a real-time serving endpoint to align with the data split used during model development?

To ensure the endpoint uses the same data for training and serving

To guarantee that the endpoint will never receive out-of-distribution data

To ensure that the serving endpoint applies the same preprocessing and feature transformations as the training pipeline

To reduce the latency of the endpoint by caching the training data

A global e-commerce company deploys a recommendation model to two endpoints: one for North America and one for Europe. The North America endpoint receives 60% of the traffic. The team notices the Europe endpoint's accuracy is degrading over time. They suspect data skew. What is the most effective way to confirm and address this?

Compare the feature distributions of the Europe serving data with the training data and retrain if significant skew is detected.

Reduce the Europe endpoint's request timeout to lower latency.

Route all Europe traffic to the North America endpoint.

Increase the Europe endpoint's compute resources to improve accuracy.

A manufacturing company deploys a predictive-maintenance model to a real-time endpoint. The training data includes sensor readings sampled every minute. At serving time, the endpoint receives data every second. What is the most appropriate configuration to align the serving data with the training data?

Aggregate the serving data to one-minute intervals before sending it to the model.

Increase the endpoint's request timeout to handle more data.

Retrain the model on one-second data.

Send the raw one-second data directly to the model.

DatabricksCertified Machine Learning Associate

Domain 4Objective 6

Split Data Between Endpoints for Realtime Interference MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 5)

Part of the Section 4: Model Deployment domain, which makes up ~30% of our current practice bank. Databricks does not publish an official question count, but from its 90-minute exam (~35–60 total, ~11–18 in this domain), expect 2–3 from this objective — we provide 23 practice questions to prepare you well beyond it. (estimate)

23questions here

5free pages

4concepts

Questions 21–23

21foundation · easy
What is the purpose of configuring a real-time serving endpoint to align with the data split used during model development?
Select an answer first
22expert · hard
A global e-commerce company deploys a recommendation model to two endpoints: one for North America and one for Europe. The North America endpoint receives 60% of the traffic. The team notices the Europe endpoint's accuracy is degrading over time. They suspect data skew. What is the most effective way to confirm and address this?
Select an answer first
23application · medium
A manufacturing company deploys a predictive-maintenance model to a real-time endpoint. The training data includes sensor readings sampled every minute. At serving time, the endpoint receives data every second. What is the most appropriate configuration to align the serving data with the training data?
Select an answer first

Finished these 3 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to MACHINE-LEARNING-ASSOCIATE

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Databricks. “MACHINE-LEARNING-ASSOCIATE” is a trademark of its owner, used for identification only.

Split Data Between Endpoints for Realtime Interference MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 5)

Questions 21–23

What is the purpose of configuring a real-time serving endpoint to align with the data split used during model development?

Keep studying — Section 4: Model Deployment