A team is clustering sensor data from industrial machines. The data contains normal operating patterns and rare fault patterns. The faults are not dense and appear as small clusters or isolated points. They used DBSCAN with epsilon=0.5 and minPts=10. The result labels most normal points as one large cluster, but many fault points are also included in that cluster. They want to separate faults as noise or separate clusters. What should they do?

Use OPTICS with a smaller minPts

Decrease epsilon and increase minPts

Increase epsilon and decrease minPts

What does a dendrogram represent in hierarchical clustering?

The hierarchy of clusters and the distances at which they merge or split

The optimal number of clusters automatically

The final cluster assignments only

A data scientist is running K-Means on a dataset and notices that the clustering results vary significantly between runs with different random initializations. They want to obtain more stable results. Which approach is most appropriate?

Use K-Means++ initialization, which selects initial centroids that are spread out, reducing variability.

Run K-Means once with a fixed random seed and use those results.

Use a different distance metric, such as cosine similarity.

Increase the number of iterations to ensure convergence.

What is a key characteristic of Gaussian Mixture Models (GMM) compared to K-Means?

GMM provides a probability of membership for each point to each cluster

GMM does not require the number of components to be specified

GMM assumes all clusters have the same variance

GMM assigns each point to exactly one cluster

How does DBSCAN handle noise points?

It assigns them to the nearest cluster

It marks them as noise and does not assign them to any cluster

It creates a separate cluster for each noise point

It removes them from the dataset before clustering

GIAC Machine Learning Engineer

Domain 2Objective 1

Clustering GMLE Practice Questions (Page 2)

Part of the Machine Learning Algorithms domain, which makes up ~29% of our current practice bank. GIAC (SANS) does not publish an official question count, but from its 180-minute exam (~70–120 total, ~20–35 in this domain), expect 7–12 from this objective — we provide 45 practice questions to prepare you well beyond it. (estimate)

45questions here

9free pages

8concepts

Questions 6–10

6expert · hard
A team is clustering sensor data from industrial machines. The data contains normal operating patterns and rare fault patterns. The faults are not dense and appear as small clusters or isolated points. They used DBSCAN with epsilon=0.5 and minPts=10. The result labels most normal points as one large cluster, but many fault points are also included in that cluster. They want to separate faults as noise or separate clusters. What should they do?
Select an answer first
7foundation · easy
What does a dendrogram represent in hierarchical clustering?
Select an answer first
8expert · hard
A data scientist is running K-Means on a dataset and notices that the clustering results vary significantly between runs with different random initializations. They want to obtain more stable results. Which approach is most appropriate?
Select an answer first
9foundation · easy
What is a key characteristic of Gaussian Mixture Models (GMM) compared to K-Means?
Select an answer first
10foundation · easy
How does DBSCAN handle noise points?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to GMLE

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 GIAC (SANS). “GMLE” is a trademark of its owner, used for identification only.