Gopher Learn – Neural networks in Golang

Gopher Learn is a library, that allows to easily train, test and implement machine learning in production environments. For this GL comes with a lot of features.

Go to Gopher Learn

Core Features

  • Training & Testing framework – GL allow to train and test, split sample sets and has options to optimize the network
  • Training using the Engine – The engine can be seen as a wrapper around the training framework. The engine helps you to find the optimally trained network.
  • Online Learning – Allows to train the network on demand and retrain it – also inject new data points – during the process

Roadmap

  • Autoencoders – With autoencoders the input data gets encoded by decisions of the network. This is useful if you are using a large float input space or strings (all sorts of ways).
  • 1vsMany – In order to scale such a network also during stream learning (online learning) 1vsMany classifiers are helpful.
  • Textos – Extensive autoencoder for large text.
  • Recos – Recommendation engine based on collaborative filtering that able to adapt and re-fit online.