This is a Neural network implementation in FORTRAN.

The size, number of hidden layers and activation function type can be selected. Settings are in the nn.inp file. See run folders for examples.

Training is done via gradient descent or via genetic optimization.