AI Engineering Degree Practice Exam 2025 - Free AI Engineering Practice Questions and Study Guide

The statement that centroids are recalculated after the initial cluster assignment is true and is a fundamental aspect of the k-means clustering algorithm. K-means starts by initializing centroids for each cluster, typically at random locations. Once initial centroids are established, the algorithm proceeds to assign each data point to the nearest centroid, forming clusters.

After the initial assignments, the algorithm recalculates the centroids based on the current clusters. This recalculation involves finding the mean position of all the data points assigned to each cluster, effectively adjusting the centroid to better represent the center of its corresponding cluster. This iterative process of assignment and centroid recalculation continues until the centroids no longer change significantly, or until a maximum number of iterations is reached, meaning the algorithm has converged.

The other statements do not correctly describe the behavior of centroids in k-means clustering. Centroids can and do move throughout the iterations, and while they represent the center of the data points in a cluster, they do not have to be located on the actual data points themselves, nor do they represent the mode of the data, but rather the average (mean).