Skip to content

MichelBartz/NNet

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NNet

Build Status

This is a C implementation of the neural network for handwriting recognition outlined in the free online book by Michael Nielsen.

It was written as a learning exercise, and is essentially a port of Michael's Python implementation.

Build instructions

  • Install the GNU Scientic library, GCC, G++

  • Download the training data:

wget http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz
wget http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz
  • Unzip the data into ./dat:
gunzip -c train-images-idx3-ubyte.gz > dat/train-images-idx3-ubyte
gunzip -c train-labels-idx1-ubyte.gz > dat/train-labels-idx1-ubyte
  • Build using cmake eg. from the project directory:

    • cd build
    • cmake ..
    • make
  • Run from the project folder:

    • Tests with ./tests
    • Train the network with ./run
  • Read the book!

About

Neural network for handwriting recognition

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • C++ 91.2%
  • C 8.6%
  • CMake 0.2%