A data scientist is training a random forest and wants to use out-of-bag (OOB) error to guide hyperparameter tuning. They have a small dataset and cannot afford a separate validation set. What is the correct way to use OOB error?

Use OOB error only for classification, not regression.

Use OOB error as a training objective to minimize during training.

Use OOB error only for feature selection, not for hyperparameter tuning.

Compare OOB error across different hyperparameter settings and choose the one with the lowest OOB error.

A data scientist has trained a random forest and wants to identify which features are most important for the model's predictions. They are concerned that impurity-based importance may be biased toward high-cardinality features. Which alternative should they use?

Tree depth at which the feature is first used.

What is the typical default number of features considered at each split for a classification random forest?

The logarithm of the total number of features

The square root of the total number of features

One-third of the total number of features

What is the purpose of bootstrap sampling in bagging for random forests?

To reduce the number of features used in each tree

To create diverse training subsets for each tree

To ensure every tree sees the same data

To eliminate the need for a validation set

What two techniques does a random forest combine to build an ensemble of decision trees?

Cross-validation and grid search

Bagging and random feature selection

Regularization and feature scaling

Dell Data Science Optimize

Domain 5Objective 2

Random Forests DATA-SCIENCE-OPTIMIZE Practice Questions (Page 3)

Part of the Data Science Theory and Methods domain, which accounts for 15% of the DATA-SCIENCE-OPTIMIZE exam.

26questions here

6free pages

8concepts

15%of the exam

Questions 11–15

11application · medium
A data scientist is training a random forest and wants to use out-of-bag (OOB) error to guide hyperparameter tuning. They have a small dataset and cannot afford a separate validation set. What is the correct way to use OOB error?
Select an answer first
12application · medium
A data scientist has trained a random forest and wants to identify which features are most important for the model's predictions. They are concerned that impurity-based importance may be biased toward high-cardinality features. Which alternative should they use?
Select an answer first
13foundation · easy
What is the typical default number of features considered at each split for a classification random forest?
Select an answer first
14foundation · easy
What is the purpose of bootstrap sampling in bagging for random forests?
Select an answer first
15foundation · easy
What two techniques does a random forest combine to build an ensemble of decision trees?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to DATA-SCIENCE-OPTIMIZE

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 Dell Technologies. “DATA-SCIENCE-OPTIMIZE” is a trademark of its owner, used for identification only.