A data scientist is comparing two models for a rare event prediction. Model A has an AUC-ROC of 0.90, and Model B has an AUC-ROC of 0.85. However, Model A has a precision of 0.02 at the desired recall of 0.80, while Model B has a precision of 0.04 at the same recall. Which model should be chosen if the business requires a precision of at least 0.03 at 80% recall?

Neither model, because AUC-ROC is not a suitable metric for imbalanced data

Model A, because it has a higher recall

Model A, because it has a higher AUC-ROC

Model B, because it meets the precision requirement at the desired recall

A team is working on a binary classification problem with a 100:1 imbalance. They have tried SMOTE and class weighting separately, but neither fully solves the problem. They decide to combine both techniques. What is the most likely benefit of this combination?

It eliminates the need for evaluation metrics

It reduces the size of the dataset

It guarantees the model will achieve perfect recall

It can provide a more balanced training set and further emphasize the minority class in the loss function

A data scientist is evaluating a model for a binary classification problem with a 1% positive class. The model's precision-recall curve shows a steep drop in precision as recall increases beyond 0.5. What does this indicate about the model's performance?

The model is overfitting the training data

The model can achieve high recall only at the cost of many false positives

The model has a high false negative rate at low recall

The model is performing better than random

A team is working on a binary classification problem with a 5% minority class. They have a small dataset of 2,000 samples. They want to use SMOTE to balance the classes. What is a potential risk of applying SMOTE to a small dataset?

SMOTE may create synthetic samples that are too similar to existing ones, leading to overfitting

SMOTE will reduce the size of the dataset

SMOTE will eliminate the need for class weighting

SMOTE will guarantee a balanced test set

DatabricksCertified Machine Learning Associate

Domain 3Objective 2

Data Preparation and Imbalance MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 2)

Part of the Section 3: Model Development domain, which makes up ~23% of our current practice bank. Databricks does not publish an official question count, but from its 90-minute exam (~35–60 total, ~8–14 in this domain), expect 2–4 from this objective — we provide 21 practice questions to prepare you well beyond it. (estimate)

21questions here

5free pages

4concepts

Questions 6–10

6expert · medium
A data scientist is comparing two models for a rare event prediction. Model A has an AUC-ROC of 0.90, and Model B has an AUC-ROC of 0.85. However, Model A has a precision of 0.02 at the desired recall of 0.80, while Model B has a precision of 0.04 at the same recall. Which model should be chosen if the business requires a precision of at least 0.03 at 80% recall?
Select an answer first
7expert · medium
A team is working on a binary classification problem with a 100:1 imbalance. They have tried SMOTE and class weighting separately, but neither fully solves the problem. They decide to combine both techniques. What is the most likely benefit of this combination?
Select an answer first
8application · medium
A data scientist is training a model to detect spam emails, where spam constitutes 20% of the emails. They want to evaluate the model's performance. Which metric would be most informative if the cost of missing a spam email (false negative) is much higher than the cost of flagging a legitimate email as spam (false positive)?
Select an answer first
9expert · medium
A data scientist is evaluating a model for a binary classification problem with a 1% positive class. The model's precision-recall curve shows a steep drop in precision as recall increases beyond 0.5. What does this indicate about the model's performance?
Select an answer first
10application · medium
A team is working on a binary classification problem with a 5% minority class. They have a small dataset of 2,000 samples. They want to use SMOTE to balance the classes. What is a potential risk of applying SMOTE to a small dataset?
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.

Data Preparation and Imbalance MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

A team is working on a binary classification problem with a 100:1 imbalance. They have tried SMOTE and class weighting separately, but neither fully solves the problem. They decide to combine both techniques. What is the most likely benefit of this combination?

A data scientist is evaluating a model for a binary classification problem with a 1% positive class. The model's precision-recall curve shows a steep drop in precision as recall increases beyond 0.5. What does this indicate about the model's performance?

A team is working on a binary classification problem with a 5% minority class. They have a small dataset of 2,000 samples. They want to use SMOTE to balance the classes. What is a potential risk of applying SMOTE to a small dataset?

Keep studying — Section 3: Model Development