In the experiment window,

1. Select the clustering method on the top right.
2. To plot data points and means on the graph - load custom datasets. or mark manually or randomly generate using the options in the side pane.
3. 'Start' will begin the clustering process.
4. 'Run' will run one iteration of the clustering process. (Applicable for K-means method).
5. 'Finish' will complete the clustering process to reach an optimal solution. (Applicable for K-means method).