In a SparkML pipeline, what is the role of a Transformer?

It evaluates the performance of a model on a test set.

It transforms a DataFrame by adding or modifying columns.

It fits a model to the training data and returns a fitted model.

It converts a DataFrame into an RDD.

A data scientist is using CrossValidator to tune a LogisticRegression model. They notice that the cross-validation results are highly variable across folds. What is the most likely cause and what should they do?

The parameter grid is too small, so they should add more parameters to tune.

The dataset is large, so they should use fewer folds (e.g., 2) to reduce variance.

The evaluator is incorrect, so they should switch to a different metric.

The dataset is small, so they should use more folds (e.g., 10 instead of 3) to reduce variance.

A data scientist is tuning a RandomForestClassifier for a binary classification problem. They have a large dataset and limited compute resources. They need to choose between CrossValidator and TrainValidationSplit. What is the most important consideration?

TrainValidationSplit is more reliable because it uses a larger validation set.

TrainValidationSplit is always better because it is faster and uses more data for training.

CrossValidator is more reliable because it uses k-fold cross-validation, but it is computationally more expensive.

CrossValidator should be used only for small datasets, while TrainValidationSplit is for large datasets.

A company needs to score a 10 TB dataset every night. Which inference approach is most appropriate?

A Structured Streaming job with a 1-second trigger

A single-node model loaded into memory and scoring row by row

A REST API that scores each record as it is generated

A SparkML model applied to the dataset in batch

DatabricksCertified Machine Learning Professional

Domain 1Objective 1

Using Spark ML MACHINE-LEARNING-PROFESSIONAL Practice Questions (Page 2)

Part of the Model Development domain, which accounts for 44% of the MACHINE-LEARNING-PROFESSIONAL exam. Databricks does not publish an official question count, but from its 120-minute exam (~50–80 total, ~22–35 in this domain), expect 6–9 from this objective — we provide 22 practice questions to prepare you well beyond it. (estimate)

22questions here

5free pages

7concepts

44%of the exam

Questions 6–10

6foundation · easy
In a SparkML pipeline, what is the role of a Transformer?
Select an answer first
7expert · hard
A data scientist is using CrossValidator to tune a LogisticRegression model. They notice that the cross-validation results are highly variable across folds. What is the most likely cause and what should they do?
Select an answer first
8expert · hard
A data scientist is tuning a RandomForestClassifier for a binary classification problem. They have a large dataset and limited compute resources. They need to choose between CrossValidator and TrainValidationSplit. What is the most important consideration?
Select an answer first
9foundation · easy
A company needs to score a 10 TB dataset every night. Which inference approach is most appropriate?
Select an answer first
10application · medium
A data engineer is building a pipeline for a regression problem. They have a numeric feature `income` that is highly skewed. They want to apply a transformation to make it more normally distributed. Which transformer should they use?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to MACHINE-LEARNING-PROFESSIONAL

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-PROFESSIONAL” is a trademark of its owner, used for identification only.