nnet is a C++ library for artificial neural networks.
It provides generic classes (neurons, synapses, networks) that can be derived
to implement specific network architectures. It is designed to be as generic as
possible, theoretically allowing to build any desired network architecture.
A derived class for simple feed-forward neural networks with a
back-propagation learning algorithm is already provided. Hopefully other
network architectures and learning algorithms will be added soon (Hebbian
Learning?).

The (still small ;o) package does include a second library for loading
and saving of neural networks.

nnet is open source and can be modified and copied under the terms
of the GNU GPL (see file COPYING).

Rosario M. Piro
