Go (golang) implementations of various Neural Networks.

• som.go is simple implementation of Self-Organizing Maps also known as Kohonen's maps.
• backprop.go is backpropagation training based neural network.

Check out demo.go for few examples on how networks can be used.

Examples in demo.go are too simplistic for use in real life. They just show how neural networks can be used. Some of examples, like prime number prediction, can be solved with simple look-up table.

Same idea applies to number of problems in the real world - if we would have all possible examples of input data we could put them into big look-up table and there would be no need for neural networks at all. Generalisation makes neural networks useful. They can produce sensible outputs for inputs that weren't encoutered during learning.