[Neurostat-develop] Some code (yes it's really happen)

[Joseph Rynkiewicz]

Yes folks, the project is really not entirely dead!

First I have to say that I agree totally with the previous project 
description of Fabrice.

I  have just import a implementation of the "C" interface beetween the 
mlp code and the linear algebra libraries (atlas, blas and a sparse blas 
library called "spblastk0.9b"). There is also  the code for the layer 
implementation following the API of the previous messages.

Comments on these files are welcome, Dimitri begin to work on the mlp 
structure. It will be good if we have the "C" version of the propagation 
and back_propagation functions for the end of June.

Joseph