Clustering

What is the primary objective of K-Means clustering?

• Maximize intra-cluster distance

• Minimize intra-cluster distance

• Maximize inter-cluster distance

• Minimize inter-cluster similarity

In K-Means, the initial centroids are chosen

• Randomly from the data points

• Using hierarchical clustering

• Based on maximum distance between points

• By minimizing sum of squares

K-Means clustering is best suited for

• Linear classification problems

• Clustering data with well-separated clusters

• Datasets with overlapping classes

• Clustering hierarchical data

How does K-Means update cluster centroids?

• By selecting the point farthest from the initial centroid

• By choosing a new random point after every iteration

• By calculating the mean of all points assigned to the cluster

• By minimizing the distance between cluster centroids

What is the stopping criterion in K-Means clustering?

• When centroids stop moving significantly

• When all points are clustered correctly

• When the number of iterations exceeds a threshold

• When intra-cluster distance is maximized

What is the primary limitation of K-Means clustering?

• Cannot handle large datasets

• Sensitive to the initial choice of centroids

• Works only on categorical data

• Requires supervised learning

If the value of K is too small in K-Means, what is the likely outcome?

• High inter-cluster variance

• Low intra-cluster distance

• Poor separation between clusters

• More accurate clustering

What does DBSCAN use to identify clusters?

• Centroid similarity

• Density of points in a region

• Distance from origin

• Hierarchical linkage

How does DBSCAN handle noise points?

• Assigns them to the nearest cluster

• Marks them as outliers

• Merges them with the smallest cluster

• Removes them from the dataset

Which of the following is an advantage of hierarchical clustering over K-Means?

• Faster computation for large datasets

• Automatically detects the optimal number of clusters

• Does not require predefining the number of clusters

• Handles noise better