A fast matrix-based approach for learning a Multi layer perceptron neural network of MNIST data using Stochastic gradient descent
The code works on visual studio and Mingw. It has MKL support and it can be build with Intel C++ compiler as well.
G++ command for Mingw
g++ -O3 -DNDEBUG -std=c++14 NNetFast.cpp -o NNetFast.exe -I"eigenSrc" \
-I"%mklroot%\include" -I"D:\Software\Software\curl\include" -L"%mklroot%\lib\intel64"\
-L"D:\Software\Software\curl\lib" -L"C:\Program Files (x86)\Windows Kits\10\Lib\10.0.14393.0\um\x64" \
-DCURL_STATICLIB -static -fopenmp -lmkl_rt -lcurl -lwldap32 -lws2_32
Neural Networks and Deep Learning http://neuralnetworksanddeeplearning.com/chap2.html via "Michael Nielson"