A data scientist is clustering a dataset that contains two overlapping Gaussian distributions. They fit a GMM with 2 components and a K-Means with k=2. The GMM provides a higher log-likelihood, but the silhouette score for K-Means is higher. Which model should they choose for downstream analysis?

Choose K-Means because the silhouette score is higher

Choose GMM because it models the data distribution better

Combine both models by averaging their assignments

Choose the model with the lower Davies-Bouldin index

A data engineer is preparing a dataset for clustering. The dataset contains a categorical feature 'region' with values 'North', 'South', 'East', 'West' and several numerical features. They plan to use K-Means. What is the best way to handle the categorical feature?

One-hot encode the categorical feature and then standardize all features before applying K-Means.

Label encode the categorical feature as 0, 1, 2, 3 and use it directly in K-Means.

Convert the categorical feature to a numerical value using the frequency of each category.

Remove the categorical feature because K-Means cannot handle categorical data.

A marketing analyst wants to segment customers based on annual income (in dollars) and number of purchases per year. They plan to use K-Means with k=4. Before running the algorithm, they ask whether they should standardize the features. What is the best advice?

Standardize both features to zero mean and unit variance so that income does not dominate the Euclidean distance.

Standardize only the income feature because purchase counts are already on a small scale.

Apply a log transformation to both features instead of standardizing.

Do not standardize; K-Means is invariant to feature scaling because it only uses relative distances.

A data scientist is comparing K-Means and DBSCAN on a dataset with two well-separated spherical clusters and some noise points. They compute the silhouette score for both clusterings. The silhouette score for K-Means is 0.8, while for DBSCAN it is 0.6. What is the most likely reason for this difference?

K-Means is better suited for spherical clusters, while DBSCAN may label noise points that lower the silhouette score

DBSCAN is not a valid clustering algorithm for this type of data

K-Means always produces higher silhouette scores than DBSCAN

The silhouette score is not a valid metric for DBSCAN

GIAC Machine Learning Engineer

Domain 2Objective 1

Clustering GMLE Practice Questions (Page 5)

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 21–25

21expert · hard
A data scientist is clustering a dataset that contains two overlapping Gaussian distributions. They fit a GMM with 2 components and a K-Means with k=2. The GMM provides a higher log-likelihood, but the silhouette score for K-Means is higher. Which model should they choose for downstream analysis?
Select an answer first
22application · medium
A data engineer is preparing a dataset for clustering. The dataset contains a categorical feature 'region' with values 'North', 'South', 'East', 'West' and several numerical features. They plan to use K-Means. What is the best way to handle the categorical feature?
Select an answer first
23application · medium
A marketing analyst wants to segment customers based on annual income (in dollars) and number of purchases per year. They plan to use K-Means with k=4. Before running the algorithm, they ask whether they should standardize the features. What is the best advice?
Select an answer first
24application · medium
A data scientist is comparing K-Means and DBSCAN on a dataset with two well-separated spherical clusters and some noise points. They compute the silhouette score for both clusterings. The silhouette score for K-Means is 0.8, while for DBSCAN it is 0.6. What is the most likely reason for this difference?
Select an answer first
25foundation · easy
Which linkage criterion in agglomerative hierarchical clustering defines the distance between two clusters as the maximum distance between any two points in the clusters?
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.

Clustering GMLE Practice Questions (Page 5)

Questions 21–25

A data scientist is clustering a dataset that contains two overlapping Gaussian distributions. They fit a GMM with 2 components and a K-Means with k=2. The GMM provides a higher log-likelihood, but the silhouette score for K-Means is higher. Which model should they choose for downstream analysis?

A data engineer is preparing a dataset for clustering. The dataset contains a categorical feature 'region' with values 'North', 'South', 'East', 'West' and several numerical features. They plan to use K-Means. What is the best way to handle the categorical feature?

A marketing analyst wants to segment customers based on annual income (in dollars) and number of purchases per year. They plan to use K-Means with k=4. Before running the algorithm, they ask whether they should standardize the features. What is the best advice?

Which linkage criterion in agglomerative hierarchical clustering defines the distance between two clusters as the maximum distance between any two points in the clusters?

Keep studying — Machine Learning Algorithms