Neural Network Software Ressources for Developers

Software for deveolpers of individual or novel applications, plugins and solutions incorporates source code, repositories of applets or components for integration in individual software programs. They are grouped by programming language.

C++

• Back Propagation Example: XOR Net
  resume: A simple C++ Neural Network class used to demonstrate a XOR task.
  description: Together with a small back propagation tutorial over on Generation 5, this simple C++ Neural Network class is easy to understand. It illustrates the use back propagation for training of a XOR task. The code compiles with MS Visual Studio, and probably other platforms with minor adjustments.
  url: http://www.generation5.org/xornet.shtml
• L.A.N.E.
  resume: A C++ library including toolkit aimed to facilitate prototyping.
  description: L.A.N.E. is a GPL library that aims to "bridge the gap between inflexible tools and rapid prototyping", according to the author. The code is available under CVS, and should work with standard C++ compilers. Toolkits are provided to ease the design of the Neural Networks.
  url: http://www.neuroquest.com/
• Kohonen Network Part 1 Tutorial
  resume: A practical approach to the Kohonen network.
  description: This tutorial takes a practical approach to the Kohonen network by applying it to a grid of 10x10, and checking if the network can cover all possible input vectors. C++ source code is provided.
  url: http://members.nbci.com/omarvision/tutorials/neuralnet/kohonen.htm
• Neural Nets C++ Library
  resume: A fully fledged C++ library, still under development.
  description: Although still under conceptual re-organisation, this library is designed in a modular fashion, and hence heavily based on virtual classes. Some of the back propagation code is not fully working. The code is written in Borland C++ builder, but the OO source is pretty much standard.
  url: http://www.cog.brown.edu/~rodrigo/neural_nets_library.html
• Neural Nets C++ Library
  author: Rodrigo de Salvo Braz
  resume: A neural networks C++ library.
  description: This library uses OO to define base classes with most of the functionality of neural nets, leaving only the essentials to be defined in new models. It already contains a backprop algorithm, as well as perceptron and winnow. In fact, it is now a part of a larger Machine Learning library that also features decision trees and naive bayes. It is under active development.
  url: http://www.cog.brown.edu/~rodrigo/neural_nets_library.html

Java

• DemoGNG
  resume: Java applet implemeting various methods to competitive learning.
  description: DemoGNG, a Java applet, implements several methods related to competitive learning. It is possible to experiment with the methods using various data distributions and observe the learning process. The source seems to work with most java implementations. A comprehensive documentation can also be found with the source archive, and online.
  url: http://www.neuroinformatik.ruhr-uni-bochum.de/ini/VDM/research/gsn/DemoGNG/DemoGNGcode.html

  Rate this resource!

• JaNet: Java Neural Network Toolkit
  resume: A well documented toolkit for designing and training, and a java library for inclusion in third party programs.
  description: jaNet package is a java neural network toolkit, which you can use to design, test, train and optimize an ideal Neural Network for your private application. You can then include your saved network in your program using the jaNet.backprop package. The consequent documentation is only in french for the moment, but an english translation is planned. The java source code is released under GPL, and can be compiled with JDK, Symantec Cafe or MS Visual J++.
  url: http://www.hta-bi.bfh.ch/Projects/janet/jaNet-1.0beta1.html
• BPN.class
  resume: A simple java class trained with backpropagation for a XOR task.
  description: The browser based java demo uses 10k training examples to perform a XOR task. The absolute error is displayed regularly to inform the user of the learning rate. The source java class is provided, and seems to work with JDK.
  url: http://www.patol.com/java/NN/index.html

Ansi C

• RCC: Recurrent Cascade-Correlation Learning Algorithm in Lisp and C
  resume: This directory contains Lisp and C implementations of the recurrent cascade-correlation learning algorithm.
  description: The source for recurrent cascade-correlation learning algorithm, implemented in C and lisp. It is released in the public domain.
  url: http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/neural/systems/rcc/0.html
• GENESIS: General Neural Simulation System
  resume: A simulation platform developed to support neural systems.
  description: This is a general purpose simulation platform which was developed to support the simulation of neural systems ranging from complex models of single neurons to simulations of large networks made up of more abstract neuronal components. Most current GENESIS applications involve realistic simulations of biological neural systems. Although the software can also model more abstract networks, other simulators are more suitable for backpropagation and similar connectionist modeling.
  url: http://www.bbb.caltech.edu/GENESIS/
• CASCOR: Lisp and C implementations of Cascade Correlation
  resume: Simple implementations of the Cascade Correlation Algorithm.
  description: A fairly old implementation for the Cascade Correlation Algorithm, with source in multiple languages. It is distributed in the public domain.
  url: http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/neural/systems/cascor/0.html
• NeurDS: Neural Design and Simulation System
  resume: A general purpose tool for building, running and analysing Neural Network Models.
  description: This is a general purpose tool for building, running and analysing Neural Network Models in an efficient manner. NeurDS will compile and run virtually any Neural Network Model using a consistent user interface that may be either window or batch oriented. The source is released under a custom licence, and should run with most standard compilers.
  url: http://hpux.u-aizu.ac.jp/hppd/hpux/NeuralNets/NeurDS-3.1/
• Neural Networks for Face Recognition
  resume: Back propagation Neural Network implemented for feature detection.
  description: This source code is released in conjunction with the Machine Learning text book. It provides an implementation of the Neural Network described in Chapter 4, where a database of facial images is used as a training set. The code is Ansi C, and should run on any compatible compiler.
  url: http://www.cs.cmu.edu/afs/cs.cmu.edu/user/mitchell/ftp/faces.html
  rating: 9
  votes: 2
• Quickprop: Backpropagation variation in Lisp, Scheme, and C
  resume: Implementation of the Quickprop algorithm.
  description: A public domain implementation of the Quickprop algorithm, in various programming languages: lisp, scheme and C.
  url: http://www.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/neural/systems/qprop/0.html
• Neural Networks at your Fingertips
  resume: Portable source for 8 popular neural network simulators.
  description: This is a collection of implementations for numerous famous simulators: Adaline, Back Propagation Network, Hopfield Model, Bi-directional Associative Memory, Boltzman, Counter Propagation Network, Self-organising Map, Adaptive Resonance Theroy. The code is portable Ansi C, and was compiled with MS Visual Studio 4.0. The source will be used in an upcoming text-book.
  url: http://www.geocities.com/CapeCanaveral/1624/