Hi Adam,

Glad you’ve enjoyed my articles!

  1. I think and RNN is probably the best way to accomplish what you have in mind. I would also recommend looking at WaveNet, https://arxiv.org/abs/1609.03499 . While that deals with generating audio, the input process would be applied to reading a continuous stream.
  2. Instead of discrete categories, it would make more sense to utilize an embedding space, where each dimension would become correlated with some aspect of the music. In this conception different clusters within the embedding space would correspond to genres, and new ones would come about naturally. This paper in particular addresses what I am talking about : http://papers.nips.cc/paper/5004-deep-content-based-music-recommendation.pdf
  3. There isn’t an explicit collection of Deep Learning models and implementations, but Github serves as a great resource. If there is any model you are interested, I would recommend doing a Github search, and odds are that someone has implemented it.

Hope those answers help!

PhD. Interests include Deep (Reinforcement) Learning, Computational Neuroscience, and Phenomenology.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store