3 ms·

what would be really awesome is if Andrew Ng or Norvig build a course around Tensorflow. It is really not useful to a beginner to be learning everything in matl

• eob 11y ago
  My first thought when reading the docs was: this is a game changer for the efficiency of applied ML (nlp, vision, speech) phd students. Coming from a school where there's a bit of a libertarian "write it all yourself, from scratch!" ethos, I always marveled at how much mileage other students got from research groups that built off a common codebase. Exciting to start to see glimmers of that possibility across the entire field.
• davmre 11y ago
  There are already a few university ML courses out there using Theano (for which Tensorflow is essentially a drop-in replacement), and I think this will be a much bigger trend over the next few years. IMHO for a first course it's useful to do some work at the Matlab/numpy level just so you get experience with deriving/implementing gradients yourself, but for larger (deep) models automatic differentiation is an amazing productivity boost that should make it possible to cover a lot of interesting topics that you'd otherwise not have space for.
  • sandGorgon 11y ago
    actually numpy would be a brilliant starting point - but I'm not able to find any popular ones that dont use matlab (or some dialect thereof). disclaimer: I have no idea what I'm talking about, but I do know that coursera and stanford courses are the oft cited ones and they use matlab/octave.
  • robert-zaremba 11y ago
    The same goes for Torch: New York, Oxford ...