K-Means Clustering

Click on canvas to add points. Use + or - to add/remove cores.

 

How does it work

Random data-points are being placed. From these points n of theme are choosed as a core.
1. Iterate over all data-points and associate them to the closest core.
2. Build the mean from all points that are associated to a core.
3. Set the cores position to the calculated mean (of its associated data-points).
4. Got to step 1 and repeat until the cores converge. (does not move anymore).