A matlab toolbox to perform Wasserstein Dictionary Learning or NMF
This repository contains a matlab implementation of the methods described in the article:
Antoine Rolet, Marco Cuturi and Gabriel Peyré. Fast Dictionary Learning with a Smoothed Wasserstein Loss. To appear in Proceedings of the 19th International Conference on Artificial Intelligence and Statistics (AISTATS) 2016, Cadiz, Spain. JMLR: W&CP volume 41. Supplementary material.
It also includes a script to reproduce the experiment in the introduction of the article step by step.
See LICENCE.txt for copyright and warranty information.
The easiest way to start using Wasserstein Dictionary Learning is to open up matlab and run all sections of the script test_wassersteinDL_mixture1d.m (html visualization)
The main function is wasserstein_DL. Type doc wasserstein_DL in a matlab terminal for usage information