Unsupervised Play: Machine Learning Toolkit for Max

Smith, Benjamin D. and Garnett, Guy E.

Proceedings of the International Conference on New Interfaces for Musical Expression

Machine learning models are useful and attractive tools forthe interactive computer musician, enabling a breadth of interfaces and instruments. With current consumer hardwareit becomes possible to run advanced machine learning algorithms in demanding performance situations, yet expertiseremains a prominent entry barrier for most would-be users.Currently available implementations predominantly employsupervised machine learning techniques, while the adaptive,self-organizing capabilities of unsupervised models are notgenerally available. We present a free, new toolbox of unsupervised machine learning algorithms implemented in Max5 to support real-time interactive music and video, aimedat the non-expert computer artist.