Question 4
Domain 3: Model DevelopmentIn a machine learning workflow, which statement best distinguishes an estimator from a transformer?
Correct answer: A
Explanation
Estimators learn parameters from data during fitting, while transformers apply a transformation to data, often using what was learned during fit. The key distinction is whether the object’s role is to learn from data, to transform data, or both. — official.txt
Why each option is right or wrong
A. An estimator fits to data to learn from it, while a transformer changes data using a transform operation.
The source material defines the distinction in terms of fitting and transforming data: estimators are associated with fit, and transformers are associated with transform. In this question, the correct contrast is that the estimator’s role is learning from data during fit, whereas the transformer’s role is applying a transformation to data.
B. An estimator only transforms data, while a transformer only evaluates model performance.
Transformers are tied to transform, and evaluators are not identified in the source material as transformers.
C. An estimator and a transformer are identical because both always perform the same operation on data.
Estimators are associated with fit, while transformers are associated with transform.
D. An estimator transforms raw data first, while a transformer fits a predictive model afterward.
Estimators are linked to fitting, and transformers are linked to transforming data.