The title is pretty self-explanatory, but what libraries do you use for minimisation (as part of gradient descent, etc.)? In any language that you use.
I was looking for a library for FORTRAN 77 but couldn't find anything simple (I couldn't get cernlib to install on my 32-bit system, and don't feel like reinstalling it to 64bit whilst I have no backup machine).
I'm considering using C++ instead but I don't have much experience with it. I had originally planned to use Python but couldn't get pybrain to work (it throws an assert error when I try to train it), I guess I could reimplement the matters with scipy/numpy though. Which should be fast enough.
[link] [comment]
