Which term refers to randomly selecting the initial centroids in centroid-based clustering?

Which term refers to randomly selecting the initial centroids in centroid-based clustering?

• A. Initialization
• B. Allocation
• C. Euclidean distance
• D. Recalculation

Answer: (A)

The term is initialization. In centroid-based clustering like k-means, the process begins by choosing starting centroids to seed the algorithm, and doing this randomly is a simple initialization method. This seed sets the starting points from which the algorithm repeatedly assigns points to the nearest centroid and then recalculates the centroids. Allocation refers to assigning each data point to the nearest centroid, not to choosing the starting points. Recalculation is the step that updates each centroid to be the mean of its assigned points. Euclidean distance is the distance metric used to determine which centroid a point is closest to. Understanding initialization helps explain why different runs can produce different clustering results and why smarter seeding methods can improve performance.

The term is initialization. In centroid-based clustering like k-means, the process begins by choosing starting centroids to seed the algorithm, and doing this randomly is a simple initialization method. This seed sets the starting points from which the algorithm repeatedly assigns points to the nearest centroid and then recalculates the centroids. Allocation refers to assigning each data point to the nearest centroid, not to choosing the starting points. Recalculation is the step that updates each centroid to be the mean of its assigned points. Euclidean distance is the distance metric used to determine which centroid a point is closest to. Understanding initialization helps explain why different runs can produce different clustering results and why smarter seeding methods can improve performance.