Certified Machine Learning ProfessionalDemo

Question

A machine learning engineer is building a recommendation system for online content. They use Databricks Feature Store to store features for training and inference.

Which unit test should they create?

A Test that Pandas DataFrame operations work as expected
B Test feature transformation functions
C Test MLflow Model Registry operations
D Test Feature Store lookup during inference

Explanation

Feature transformation functions contain the application-specific logic that computes feature values, so unit tests should verify that they produce the expected results for known inputs. Databricks documents unit testing feature-update code by mocking Feature Store client calls while testing the feature-computation behavior.

Learn more

Databricks Feature Engineering Python API — Use the clients for unit testing

Certified Machine Learning ProfessionalDemo

AstroTutor

Notes

Readiness

QuestionQ163

Community Discussion