This library is a C++ template-based library that provides multilayer perceptrons to gaml. It basically provides templates for creating a multilayer networks with various transfer functions (you can provide your own) which can fit to your input and output data. Learning the predictors can be done through backpropagation or Unscented Kalman Filters.
Download and documentation
The latest gaml-mlp package can be found here as a gaml-mlp-*.tar.gz package.
The documentation (with examples) is here.
Various examples for classification and regression using backpropagation or kalman filters for learning are provided in the library, check out the documentation of the examples.