'Actio Potentialis' is a small and simple artificial neural network written in C++ for Vibrowtion.
Actio Potentialis (from the Latin agō and the Late Latin potentialis) is sort of a wordplay on the premise of the 'Vibrowtion' project and the actual physiological process that allows neurons to communicate in (artificial) neural networks.
The code used for Actio Potentialis is predominantly based on the code of George Chousos. A huge thanks to them for making this project a lot easier. More in-depth information about the backpropagation algorithm is made available by Matt Mazur at A Step by Step Backpropagation Example.
Actio Potentialis is licensed under the GNU Lesser General Public License V3.0.