anthonymorast / yall

Yet Another Learning Library (I thought the name up on my own but it's not very original)

YALL (Yet Another Learning Library)

YALL is a C++ library that implements many machine learning algorithms and attempts not to use company or platform specific libraries. That is, the goal of YALL is to be readily usable on Windows, Linux, Mac, Raspberry Pi, etc. without a lot of installation and headaches.

There are many great machine learning libraries out there that do many things, probably more efficiently than YALL. The primary purpose of this library is for my own personal experience with a secondary purpose of being portable, useful, and efficient

An additional FYI, this is the first project I'm building using CMAKE. It's also acting as an introduction to this build system and enhancing my knowledge of linux linking/compiling. Excuse the note files all over the place...

Dependencies

• C++ DataTables
  • Install instructions and repository: http://www.github.com/anthonymorast/DataTables
• Armadillo (built with 9.860.1)
  • Depends on (Ubuntu versions): libopenblas-dev liblapack-dev libarpack2-dev libsuperlu-dev
  • Install Instructions for all Platforms: http://arma.sourceforge.net/download.html
• CMAKE
• Gnuplot
  • http://www.gnuplot.info/
  • Installing on Linux: should be available in whatever package manager is used for your distribution as 'gnuplot'
  • Installing on Windows
    • https://stackoverflow.com/questions/47453376/how-to-install-gnuplot-on-windows
• C++11
  • The library might work with older versions of C++ but I'm compiling my tests with --std=c++11

IN PROGRESS

Save/Load Neural Network Weights

File Structure

Custom Neural Network Activation Functions

Custom Neural Network Back Propagation Techniques