An iterative method of cluster analysis that groups large data sets into clusters of similar data. A k‑means method forms clusters around data values having the nearest mean.