Interactive K-means

Visualize and interact with the clustering algorithm k-means.
Try it at lettier.com/kmeans.
Read more about k-means.

Download & Run

git clone https://github.com/lettier/interactivekmeans.git
cd interactivekmeans
nohup python -m http.server &> /dev/null &
python -mwebbrowser http://localhost:8000

Directions

• Lay down data points by clicking the mouse.
• You can also use the scatter button located in the controls.
• Set your value for k and maxIterations.
• Press runKMeans to cluster the on-screen data points.
• Remove data points by clicking on them.
• Use the silhouette coefficient to find the optimal k.

(C) 2015 David Lettier.
lettier.com