Well here it is the code that others regarded as unholy.
My first attempt at a Restricted Boltzmann
I hope you like it, it still needs a lot more work to get it to a reasonable standard:
Faults are that it is set at only 1 gibbs sample - doesnt seem right.
Innaccurate not 100% on recall / reconstruction of Xor
Ideal for anyone wishing to code an RBM in C++ for beginers like me!
Just needs g++ and your ready to go.
Thanks