Home

Introduction

While Synaptic 2.0 is being developed, you might want to use Gynaptic in the meanwhile. It offers extra functions and genetic algorithms, as well as overall code improvement.

Welcome to Synaptic's wiki! There is tons of material on how to create (and understand!) neural networks. The library includes a few pre-built network architectures like Perceptrons and Hopfields. Those are discussed at the architecture page. If you want to roll on your own, take a look at the basics of creating and modifying a network at the network page. If you want to go even deeper, check out the layer and neuron pages. And last but not least, setting up a training scheme for your network is explained on the trainer page.

Useful links

• If you have no prior knowlegde about neural networks, start here
• If you want to know how you feed custom data to a neural network, check out this
• If you want to get started on neural networks in JS in general, read this artickle
• If you want to apply genetic algorithms on neural networks, check out this library

Demos

• Solve an XOR
• Discrete Sequence Recall Task
• Learn Image Filters
• Paint an Image
• Self Organizing Map
• Read from Wikipedia

Talks

• Machine Learning with Node.js - An introduction to neural networks on the basis of synaptic.
• Also covering how to check the performance of the algorithm and how to tackle the error rate. PDF slides