K-Means Clustering Simulation

• 

• K-means clustering might be the most popular and simplest one among the clustering methods The procedure is as follows:

  1. Specify the number of clusters.
  2. Add an initial center point for each cluster.
  3. For each data points, calculate its distance from cluster centers and choose the nearest one as the cluster of that point.
  4. Update the cluster center positions according to the points in clusters.
  5. Repeat 3-4 until the algorithm converges to a certain solution.

  I prepared the widget below long time ago. It shows the iterations of that k-means clustering process for a randomly generated data with two variables.

  • "New Data" button generates data points to use.
  • "Generate Groups" button creates random cluster centers.
  • "Iterate Groups" button does the step 3 and step 4 in the algorithm. You can see the further iterations by clicking this button.

  New Data Generate Groups Iterate Groups

machine-learning