A machine learning team is developing a model to predict equipment failure, which occurs in only 0.1% of cases. They trained a model that achieves 99.9% accuracy. Which of the following is the most important reason why accuracy is not a suitable metric here?

Accuracy does not account for the cost of false negatives

Accuracy is only applicable to regression problems

Accuracy cannot be computed for imbalanced datasets

Accuracy is always lower than F1-score

In a binary classification problem with class weights, how does assigning a higher weight to the minority class affect the loss function?

It decreases the penalty for misclassifying majority class samples.

It has no effect on the loss function but changes the learning rate.

It increases the penalty for misclassifying minority class samples.

It changes the model's architecture to handle imbalance.

A data scientist is training a random forest classifier on an imbalanced dataset. They want to use class weights to improve minority class recall. Which of the following is the most appropriate way to set class weights in scikit-learn?

Set class_weight='balanced' in the RandomForestClassifier constructor

Manually duplicate minority class samples in the training set

Set the max_depth parameter to a higher value

A machine learning engineer is training a logistic regression model on a dataset where the positive class represents a rare disease (1% prevalence). They want to penalize misclassifications of the positive class more heavily without changing the dataset. What should they do?

Apply class weights to the loss function

Switch to a decision tree algorithm

Increase the regularization parameter

DatabricksCertified Machine Learning Associate

Domain 3Objective 2

Data Preparation and Imbalance MACHINE-LEARNING-ASSOCIATE Practice Questions (Page 1)

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 1–5

1application · medium
A machine learning team is developing a model to predict equipment failure, which occurs in only 0.1% of cases. They trained a model that achieves 99.9% accuracy. Which of the following is the most important reason why accuracy is not a suitable metric here?
Select an answer first
2foundation · easy
In a binary classification problem with class weights, how does assigning a higher weight to the minority class affect the loss function?
Select an answer first
3application · medium
A data scientist is training a random forest classifier on an imbalanced dataset. They want to use class weights to improve minority class recall. Which of the following is the most appropriate way to set class weights in scikit-learn?
Select an answer first
4application · medium
A data scientist is training a binary classifier to detect fraudulent transactions. Only 0.5% of transactions are fraudulent. The initial model achieves 99.5% accuracy on the test set, but the data scientist knows this metric is misleading. Which evaluation metric would best reveal the model's ability to identify fraudulent transactions?
Select an answer first
5application · medium
A machine learning engineer is training a logistic regression model on a dataset where the positive class represents a rare disease (1% prevalence). They want to penalize misclassifications of the positive class more heavily without changing the dataset. What should they do?
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.