K-medoids Clustering Quiz

What is the primary goal of partitioning clustering?

• Maximize similarity between different clusters

• Minimize the number of clusters

• Maximize similarity within each cluster

• Minimize the distance between all data points

Which of the following is NOT a popular partitioning method?

• k-means clustering

• Hierarchical clustering

• k-medoids clustering

• CLARA clustering

In k-medoids clustering, what does a 'medoid' represent?

• The mean of the cluster

• A random data point

• An actual data point in the cluster

• The centroid of the cluster

What is the first step in the k-medoids clustering process?

• Randomly select k data points as medoids

• Compute the total cost of clustering

• Update the medoids

• Assign points to the closest medoid

Which distance metric is commonly used in k-medoids clustering?

• Euclidean distance

• Cosine similarity

• Manhattan distance

• Hamming distance

What is a disadvantage of k-medoids clustering?

• Less interpretability

• Higher computational complexity

• Robustness to outliers

• Applicability to arbitrary distance metrics

How is the total cost of clustering computed in k-medoids?

• By minimizing the distances between clusters

• By averaging the distances within clusters

• By summing the distances to the medoids

• By summing the distances between all points

What does the term 'swap cost' refer to in k-medoids clustering?

• The cost of updating medoids

• The cost of assigning points to clusters

• The cost of clustering with non-medoids

• The cost of swapping two medoids

Which of the following is an advantage of k-medoids over k-means?

• Not affected by extreme values

• Easier to implement

• More sensitive to outliers

• Less computationally intensive

What is the formula used to compute the swap cost?

• Ch = d(Oh, Oi) - d(Oh, Oj)

• Ch = d(Oh, Oj) + d(Oh, Oi)

• Ch = d(Oh, Oj) - d(Oh, Oi)

• Ch = d(Oh, Oi) + d(Oh, Oj)