What is a key difference between online and offline feature tables?

Online tables are only available in Unity Catalog, while offline tables are not.

Online tables store features in Delta Lake, while offline tables store them in a key-value store.

Online tables are used for low-latency inference, while offline tables are used for batch training and analysis.

Online tables are read-only, while offline tables can be written to.

A data scientist is training a model and wants to ensure that the same feature definitions are used during training and inference. They plan to use the Feature Store to log the model. What is the primary benefit of using the Feature Store's model registry integration?

It stores the model's training code and dependencies.

It converts the model to a format that runs on any cloud provider.

It records the feature table versions and transformations used, enabling consistent feature lookup during inference.

It automatically tunes hyperparameters for the model.

A machine learning team has trained a model using features from a feature store. During inference, they need to retrieve features for a batch of new customers. The features are stored in an offline table. Which approach should they use to score the batch?

Use a REST API call to the online feature store for each customer.

Query the feature table using Spark SQL directly.

Load the entire feature table into memory and filter for the new customers.

Use fs.read_table to load the feature table and join with the new customer IDs.

In the Databricks Feature Store, what is the purpose of defining a primary key when creating a feature table?

To determine the physical storage location of the table in cloud storage.

To enforce data quality rules on the feature values.

To uniquely identify each row and enable point lookups and joins during training and inference.

To specify which columns are used for partitioning the table.

A data scientist is training a churn prediction model. They need to load features from a feature store table named 'customer_churn_features' that resides in the 'analytics' catalog and 'ml' schema. The training DataFrame should include the label column 'churn' from a separate table. Which code snippet correctly loads the feature DataFrame for training?

training_df = spark.table('analytics.ml.customer_churn_features').join(labels_df, on='customer_id')

feature_df = fs.read_table(name='analytics.ml.customer_churn_features'); training_df = feature_df.join(labels_df, on='customer_id')

training_df = fs.read_table(name='customer_churn_features').join(labels_df, on='customer_id')

training_df = fs.read_table(name='analytics.ml.customer_churn_features')

DatabricksCertified Machine Learning Associate

Domain 1Objective 2

Feature Store Management MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 2)

Part of the Section 1: Databricks Machine Learning domain, which makes up ~19% of our current practice bank. Databricks does not publish an official question count, but from its 90-minute exam (~35–60 total, ~7–11 in this domain), expect 2–3 from this objective — we provide 18 practice questions to prepare you well beyond it. (estimate)

18questions here

4free pages

6concepts

Questions 6–10

6foundation · easy
What is a key difference between online and offline feature tables?
Select an answer first
7application · medium
A data scientist is training a model and wants to ensure that the same feature definitions are used during training and inference. They plan to use the Feature Store to log the model. What is the primary benefit of using the Feature Store's model registry integration?
Select an answer first
8expert · medium
A machine learning team has trained a model using features from a feature store. During inference, they need to retrieve features for a batch of new customers. The features are stored in an offline table. Which approach should they use to score the batch?
Select an answer first
9foundation · easy
In the Databricks Feature Store, what is the purpose of defining a primary key when creating a feature table?
Select an answer first
10application · medium
A data scientist is training a churn prediction model. They need to load features from a feature store table named 'customer_churn_features' that resides in the 'analytics' catalog and 'ml' schema. The training DataFrame should include the label column 'churn' from a separate table. Which code snippet correctly loads the feature DataFrame for training?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back 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.