k-means
Splitting points into k groups by repeatedly assigning each to the nearest centre and moving each centre to the mean of what it captured.
What people get wrong
k is chosen by you, not found by the algorithm, and it always returns exactly k groups whether or not the data has any. It also assumes roughly spherical, similarly sized clusters.
A definition is not an understanding. This is taught properly in Clustering, hands on, part of Unsupervised learning.
Learn this properlyTerms from the same topic