1.Initialize – Pick k random cluster centers
–Pick centers near data.
Heuristics: uniform distribution
in range of data; randomly select data
points.
2.Assign each point to nearest center.
3.Make each center average of pts assigned to it.
4.Go to step 2.