Skip to content

meyerscetbon/LOT

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

47 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Low-Rank Sinkhorn Factorization

Code of the paper by Meyer Scetbon, Marco Cuturi and Gabriel Peyré.

A New Regularization Scheme for computing Efficiently the Optimal Transport Cost

In this work, we propose to regularize the optimal transport (OT) problem by adding a low-rank constraint on the couplings. In the following figure, we compare the couplings obtained by the Sinkhorn algorithm (bottom row) and by our proposed scheme (top tow). figure

We propose an efficient algorithm to solve the optimal transport problem under low-rank constraints and show on several examples that our method outperforms the Sinkhorn algorithm in term of time-accuracy tradeoff. In the following figure we illustrate the main differences between our proposed algorithm and the Sinkhorn one. figure

Our regularization can take advantage of the geometry of the problem, in particular when the cost matrix involved in the optimal transport problem admits a low-rank factorization. In this case, our method is able to compute the OT cost in linear time with respect to the number of samples. We present the time-accuracy tradeoff between different methods to compute the OT when the samples are drawn from two Gaussian distributions evaluated on 5000 points in 2D and the underlying cost is the squared Euclidean distance. figure

This repository contains a Python implementation of the algorithms presented in the paper.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages