NEUGO

Overview

NEUGO is a simple neural networks framework in Go, which allows the users to easily customize a neural network by adding/removing nodes (neurons) and connections, therefore providing intuitive interfaces and flexible topologies. The framework is still in development.

Algorithms

Following are available algorithms currently. More will be added with further updates.

• Perceptron (single-layer network) (in progress)
• Feedforward neural network (2-layer network)
• MLP (Multi-layer Perceptron) (in progress)
• Hopfield Network (in progress)

Planned Algorithms

These are algorithms that are not implemented yet, but are planned to be added in the future.

• Recurrent Neural Network / LSTM Network
• Convolutional Neural Network
• Neural Turing Machine

Installation

go get github.com/jinseokYeom/neugo

Note

The purpose of this package is for me to learn more about various kinds of neural networks and their learning algorithms; I would appreciate any constructive feedbacks to improve this project!