Non-negative latent space models

This repository contains the material used for recreating the code used in: -[Non-negative matrix factorization with gaussian process priors. Computational Intelligence and Neuroscience, 2008] -[Bayesian Inference for Nonnegative Matrix Factorisation Models.Computational Intelligence and Neuroscience, 2009] Both methods outperform the standard NMF algorithm from DD- Lee 2001:

• Algorithms for Non-negative Matrix Factorization

Several files can be found including:

• Report with results of the people collaborating on the project and
• A pdf file containing attemped proofs of several statements in the mentioned papers
• Folder with code used for training and creatinf results

see our paper for more details.

The work presented in this repository is to be considered open source under the MIT License. If you found this code useful in your research, then please cite

@misc{ebert-majidi-Hatting-Refsgaard,
  title={Variants of NonNegative Matrix Factorization},
  author={Ebert, Peter, Majidi, Oldouz, Hatting, Martin, Refsgaard, Nicklas},
  year={2018}
}