Java Neural Network

A neural network is a mathematical construct that attempts to emulate the way that actual neural networks, in the brain, work. A neural network accepts input and produces output. The input and output are usually both floating point arrays. Neural networks can be implemented in a variety of programming languages. This page focuses on using neural networks with Java.

If you are a C#/DotNet programmer and would like to add neural network functionality to a C# program you should consider the Encog Artificial Intelligence Framework. This framework gives you all of the tools you need to create a variety of neural networks. Some of the neural network types supported by Encog are listed here:

• Feedforward Neural Network
• Hopfield Neural Network
• Self Organizing Map (SOM)
• Recurrent Self Organizing Map (RSOM)
• Elman Neural Network
• Jordan Neural Network
• Boltzmann Machine
• Bidirectional Associative Memory (BAM)
• ADALINE Neural Network
• Adaptive Resonance Theory (ART)
• Counterpropagation Network (CPN)

For more information about Encog, and to see C# examples of the above neural network types, visit the Encog homepage.