Distance Based Linkage Methods Quiz

In single linkage, the distance between two clusters is defined as

• The maximum distance between any pair of points from the two clusters

• The minimum distance between any pair of points from the two clusters

• The average distance between all pairs of points from the two clusters

• The distance between the centroids of the two clusters

In complete linkage, the distance between two clusters is defined as:

• The maximum distance between any pair of points from the two clusters

• The minimum distance between any pair of points from the two clusters

• The average distance between all pairs of points from the two clusters

• The distance between the centroids of the two clusters

Average linkage is also known as:

• Single-linkage

• Complete-linkage

• Group average linkage

• Ward's method

What is the primary characteristic of centroid linkage in clustering?

• The distance is based on the furthest points in the clusters

• The distance is based on the closest points in the clusters

• The distance is calculated using the centroids of the clusters

• The distance is the average of all points in the clusters

Which of the following methods is NOT a hierarchical clustering method?

• Single linkage

• Complete linkage

• K-means clustering

• Average linkage

In hierarchical clustering, what does the term 'dendrogram' refer to?

• A graphical representation of the clustering process

• A method for calculating distances between clusters

• A type of clustering algorithm

• A statistical measure of cluster validity

What is the main advantage of using hierarchical clustering over other clustering methods?

• It does not require the number of clusters to be specified in advance

• It is faster than K-means clustering

• It can handle large datasets efficiently

• It guarantees optimal clustering results

Which of the following is a common method for determining the optimal number of clusters in hierarchical clustering?

• Elbow method

• Silhouette analysis

• Gap statistic

• All of the above

In which scenario is Ward's method particularly useful in hierarchical clustering?

• When clusters are of different sizes

• When the goal is to minimize the total within-cluster variance

• When the data is not normally distributed

• When the number of clusters is known

What is the main disadvantage of using single linkage in hierarchical clustering?

• It can create long, chain-like clusters

• It is computationally expensive

• It requires the number of clusters to be specified

• It is sensitive to outliers