[Happycoders - Association]

Designed by Initd.org

Generic C++ Neural Networks library

Description

Happycoders Neural Networks C++ library provides various algorithms for MLP (Multi Layers Perceptron):

Backprop
Rprop
Quickprop

It contains principal component analysis algorithms :

ALA
GHA

It supports normal and time lagged neurons. This library is released under the term of the GNU LGPL licence.

Features

Learnings algorithms : Backprop, Rprop and Quickprop
Principal Component Analysis algorithms : ALA and GHA
Normal and time lagged neurons
file format to store neural networks structure
batch and stochastic mode
can store result in pgm picture

Installation

get last release : libneuralnet-0.1.tar.gz
get the debian package : happycoders-libneuralnet and happycoders-libneuralnet-dev add this line to /etc/apt/sources.list :
```
    deb http://debian.speedblue.org ./
```

Example

Nncompress is a application of libneuralnet, it can compress/decompress images using a MLP,

get last release : nncompress-0.42.tar.gz
get the debian package : happycoders-nncompress add this line to /etc/apt/sources.list :
```
    deb http://debian.speedblue.org ./
```

screenshot

(c) 2002, 2003, 2004, 2005, 2006 Happycoders