libSVM using Python

Welcome to /learn/. This is my play-around toy with libSVM powered by a Javascript/AJAX interface with a Django web server backend.

By choosing a label color, you can add a point to my dataset. Each point you add is sent via AJAX to a database. After adding points, train the SVM on the dataset.

Then click predict all to see the classifier in action. Press "clear" to remove the prediction to add more points and train/predict again. Pressing predict instead of predict all makes 144 separate AJAX requests to get a separate prediction at each prediction point. The loading effect is cool but if you can, please spare my bandwidth :)

Points Loaded: 164
Click to Add point