MS-DS Master of Data science FREE MS-DS Master of Data science Unsupervised Machine Learning Models Questions and Answers 3

Question 1

What problem does the elbow method help solve when using K-Means clustering?

Accepted Answer

Selecting the optimal number of clusters

Answer

Choosing the right distance metric

Answer

Determining feature importance

Answer

Selecting the initialization method

Question 2

In spectral clustering, what is the primary role of the graph Laplacian matrix?

Accepted Answer

It captures the connectivity structure of the data for eigen-decomposition

Answer

It normalizes the input features to unit variance

Answer

It computes the distance between all pairs of centroids

Answer

It performs dimensionality reduction via SVD

Question 3

Which unsupervised method is specifically designed to identify data points that do not conform to expected patterns?

Accepted Answer

Isolation Forest

Answer

K-Means clustering

Answer

Latent Dirichlet Allocation

Answer

Non-negative Matrix Factorization

Question 4

What does the Bayesian Information Criterion (BIC) balance when selecting the number of components in a Gaussian Mixture Model?

Accepted Answer

Model likelihood and model complexity

Answer

Training speed and accuracy

Answer

Feature count and sample size

Answer

Cluster diameter and density

Question 5

In the context of association rule mining, what does the 'lift' metric indicate?

Accepted Answer

The ratio of observed support to expected support if items were independent

Answer

The total number of transactions containing the itemset

Answer

The confidence of the rule divided by the number of items

Answer

The minimum support threshold for frequent itemsets

Question 6

What is the primary objective of the Self-Organizing Map (SOM) algorithm?

Accepted Answer

To project high-dimensional data onto a low-dimensional grid while preserving topological relationships

Answer

To maximize the variance of projected data along orthogonal axes

Answer

To partition data into exactly K non-overlapping clusters

Answer

To build a probabilistic generative model of the data