Free Practice Quiz Question List

6 Clustering Online Quiz Questions

Use this free practice quiz with 20 questions to review 6 Clustering, test your knowledge, and prepare for your next test or exam.

20 questions

True or false

1 point

A model is given examples with correct group labels in advance and learns to assign new examples to those labels. Is this clustering?

A
True
B
False

Choose one

1 point

A store has customer data on visit frequency and average purchase amount, but no predefined customer-group labels. Which approach best fits the goal of discovering groups with similar shopping patterns?

A
Train a classifier using the customers' predefined shopping-pattern labels.
B
Cluster customers using features such as visit frequency and average purchase amount.
C
Assign every customer to a group at random.
D
Sort customers by name and treat each name range as a group.

Fill in the blank

1 point

K-means assigns each example to the nearest of its centers.

Written response

1 point

In distance-based clustering, features with larger numerical scales can dominate distance calculations. Name one preprocessing technique that can help address this.

Choose one

1 point

A dataset appears to contain dense groups with irregular shapes and isolated examples that should be treated as noise. Which method is the best match among these options?

A
K-means, because it estimates each example's probability of belonging to every cluster.
B
Hierarchical clustering, because it must mark isolated examples as noise.
C
DBSCAN, because it can find irregularly shaped clusters and identify isolated examples as noise.
D
Gaussian mixture models, because they repeatedly merge the closest groups.

Fill in the blank

1 point

In agglomerative hierarchical clustering, each example starts alone, and the method repeatedly the closest groups.

True or false

1 point

A hierarchical clustering tree can be cut at different levels to produce different numbers of clusters.

A
True
B
False

Choose one

1 point

An analyst wants a clustering method that can express uncertainty by estimating how likely an example is to belong to each group. Which method best fits this need?

A
DBSCAN, because it builds a tree that can be cut at different levels.
B
Gaussian mixture models, because they can estimate probabilities of membership in different clusters.
C
K-means, because it marks isolated examples as noise.
D
Agglomerative hierarchical clustering, because it assigns probabilities to every cluster.

Choose all

1 point

Select all statements that accurately describe how clustering methods differ.

A
Different algorithms may define clusters in different ways and produce different groupings from the same data.
B
All clustering algorithms use the same definition of a cluster, so their results should match.
C
Methods rely on different assumptions and may suit different data shapes.
D
No single clustering method is best for every dataset.

Written response

1 point

Which density-based clustering algorithm has results that depend on settings such as neighborhood radius and minimum neighborhood size?

Choose all

1 point

A team is deciding whether discovered customer clusters are useful. Select all checks that would support a responsible interpretation of the results.

A
Check whether the groups are useful for the intended task.
B
Check whether the results change substantially when settings change.
C
Assume each cluster is a naturally existing category because the algorithm produced it.
D
Inspect whether the examples grouped together make sense.

Choose one

1 point

A researcher wants to explore a collection of unlabeled documents for groups of related topics. Which plan best uses clustering while accounting for the limits of the results?

A
Treat each resulting group as a definitive topic category without checking its contents.
B
Use clustering only after assigning every document a correct topic label.
C
Convert documents to feature vectors and cluster them to help explore related topics, then review the groups.
D
Cluster documents by their filenames, since feature representations are not useful for text.

Choose one

1 point

A dataset is expected to have a chosen number of compact, roughly round groups, and an efficient method is preferred. Which algorithm is the most suitable starting choice, while still requiring checks for issues such as outliers or starting centers?

A
DBSCAN, because it requires choosing the number of centers before clustering.
B
K-means, because the expected groups are compact and roughly round and the number of clusters is known.
C
Gaussian mixture models, because they always find irregular clusters without parameter choices.
D
Hierarchical clustering, because it cannot produce different numbers of clusters.

Open ended

1 point

You are exploring unlabeled data that appears to contain dense, irregularly shaped groups as well as isolated examples. Recommend a clustering method and explain why it fits, one setting-related limitation to examine, and how you would judge whether the resulting groups are useful.

Choose one

1 point

A team has customer data but no existing customer categories. It wants an algorithm to group customers with similar visit frequency and purchase amounts. Which task best fits this goal?

A
A model learns to assign examples to predefined labels.
B
An algorithm groups examples by feature similarity without predefined labels.
C
A model estimates the probability that a labeled example is correct.
D
An algorithm removes features that have large numerical scales.

Choose one

1 point

A distance-based clustering analysis uses two numerical features, one measured in single digits and the other in thousands. What is a sensible step to reduce the risk that the larger-scale feature dominates the distances?

A
Scale or transform features so larger numerical scales do not dominate distance calculations.
B
Choose a larger number of clusters so every feature has equal influence.
C
Use classification labels to correct the distances.
D
Remove every feature whose values differ between examples.

Choose one

1 point

Which description correctly identifies the method that repeatedly assigns examples to their nearest center and updates the centers, and states an important choice it requires?

A
It merges the closest groups repeatedly, starting with one group per example.
B
It groups examples in dense regions and labels isolated examples as noise.
C
It is K-means: it assigns examples to the nearest of k centers, updates the centers, and requires choosing k.
D
It estimates a probability distribution for each predefined label.

True or false

1 point

In agglomerative hierarchical clustering, each example begins in its own group, and the closest groups are repeatedly merged.

A
True
B
False

Written response

1 point

Which clustering method can estimate an example's probability of belonging to each cluster rather than assigning it only to one group?

Written response

1 point

When known labels are unavailable, what named measure may be used alongside practical review to evaluate clustering results?