A simple implementation of support vector clustering in only python/numpy
This implements a version of support vector clustering from the paper: "A Support Vector Method for Clustering", A. Ben-Hur et al.
The implementation is reasonably performant, and should easily handle 10-50k training sets without incremental training. I am only aware of a few other implementations of support vector clustering, none of which are simplified to this extent. Hopefully somebody finds it useful, or at least informative.