AI Paper Atlas

Research map / Learning theory and algorithms / Optimal transport

A Combinatorial Algorithm for Approximating the Optimal Transport in the Parallel and MPC Settings

Nathaniel Lahn, Sharath Raghvendra, Kaiyi Zhang

NeurIPS 2023 · Learning theory and algorithms › Optimal transport · approach: Sinkhorn and entropic OT algorithms · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
replace Sinkhorn (POT, GPU) with the parallel combinatorial algorithmOptimal transport solver · Synthetic assignment, n=10,000GPU running timeImproved
replace Sinkhorn with the parallel combinatorial algorithmOptimal transport solver · Synthetic OT with random supplies/demands, n=10,000GPU running timeImproved
replace Sinkhorn with the parallel combinatorial algorithmOptimal transport solver · MNIST assignment, n=10,000 imagesGPU running timeImproved
replace Sinkhorn with the parallel combinatorial algorithmOptimal transport solver · Text OT: Count of Monte Cristo, 20News, IMDB (GloVe)GPU running timeImproved
replace Sinkhorn with the parallel combinatorial algorithmOptimal transport solver · Synthetic, MNIST and text OTnumber of parallel roundsImproved
compare against a CUDA-based GPU implementation in addition to PyTorchParallel combinatorial algorithm · GPU implementation comparisonrunning timeImproved
match the regularization of Sinkhorn to the cost of the algorithm versus reversed (fix Sinkhorn, search epsilon)Experimental protocol · All four settingsrunning time and roundsNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy