code for neuro-symbolic integration, following Artur S. d'Avila Garcez, Krysia Broda and Dov M. Gabbay. Neural-Symbolic Learning Systems: Foundations and Applications, Perspectives in Neural Computing, Springer-Verlag, ISBN 1-85233-512-2, 2002.
The code is based on Matlab/Octave Code for Artificial Neural Networks
Simple neural network code written in pure Octave/Matlab inspired by Andrew Ng's Machine Learning Course on Coursera.
The code for training a neural network is surprisingly concise, only a dozen lines of code or so (see nn_train.m and nn_predict.m).
To get octave on MacOS, use brew install octave.
To run the demos,
$ octave demo1.m
$ octave demo2.m
$ octave demo_cilp.m