DSE - Data Science Unsupervised Learning: Clustering Questions and Answers

Question 1

A data scientist is working with a dataset containing customer transactions for an e-commerce platform.
The goal is to segment customers into distinct groups based on their purchasing behavior to enable targeted marketing campaigns.
The data contains non-globular cluster shapes and a significant amount of noise from fraudulent or one-time buyers.
The number of customer segments is not known beforehand.
Which clustering algorithm would be most appropriate for this scenario?

Accepted Answer

DBSCAN

Answer

K-Means

Answer

Agglomerative Hierarchical Clustering

Answer

Gaussian Mixture Models

Question 2

Which of the following is a significant limitation of the K-Means clustering algorithm?

Accepted Answer

It requires the user to manually specify the number of clusters, 'k', beforehand.

Answer

It is computationally very expensive, with complexity that scales exponentially with the number of data points.

Answer

It can only be applied to datasets with a small number of features (low dimensionality).

Answer

The resulting clusters are often difficult to interpret and visualize.

Question 3

In hierarchical clustering, what is the fundamental difference between the agglomerative and divisive approaches?

Accepted Answer

Agglomerative is a bottom-up approach that starts with each data point as an individual cluster and progressively merges them, while divisive is a top-down approach that starts with all points in one cluster and recursively splits them.

Answer

Agglomerative is a top-down approach, while divisive is a bottom-up approach.

Answer

Agglomerative starts with all data points in a single cluster and recursively splits them, while divisive starts with each point as its own cluster and merges them.

Answer

Agglomerative is only suitable for numerical data, while divisive can handle categorical data.

Question 4

A data scientist has just performed clustering on a dataset and wants to evaluate the quality of the resulting clusters without using any external labels. They choose to use the Davies-Bouldin Index. What would a low Davies-Bouldin Index score indicate?

Accepted Answer

Good clustering, with clusters that are compact and well-separated from each other.

Answer

Poor clustering, with high similarity between clusters.

Answer

Good clustering, with low intra-cluster similarity and high inter-cluster similarity.

Answer

The algorithm failed to converge.

Question 5

What are the two primary parameters that need to be defined when using the DBSCAN algorithm?

Accepted Answer

Epsilon (ε) and the minimum number of points (MinPts).

Answer

The number of clusters (k) and the linkage criteria.

Answer

The initial centroid locations and the number of iterations.

Answer

The distance metric and the desired dendrogram height.

Question 6

Which of the following statements best describes unsupervised learning?

Accepted Answer

A type of machine learning where the algorithm learns to find patterns and inherent structures in unlabeled data.

Answer

A type of machine learning where the algorithm learns from data that has been manually labeled with the correct output.

Answer

A type of machine learning focused on learning a mapping function from input variables to a continuous output variable.

Answer

A method of learning that involves an agent interacting with an environment to maximize a cumulative reward.