Research map / Learning theory and algorithms / Optimal transport
Near-linear time approximation algorithms for optimal transport via Sinkhorn iteration
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use greedy single-row or single-column updates instead of alternating full row and column updates | Sinkhorn projection · MNIST image pairs | distance to transport polytope | Helped |
| Use greedy single-row or single-column updates instead of alternating full row and column updates | Sinkhorn projection · Random images | distance to transport polytope | Helped |
| Use greedy updates that prioritize salient rows and columns | Sinkhorn projection · Random images with more background pixels | runtime | Helped |
| Use Greenkhorn rather than Sinkhorn at small regularization parameter values | ApproxOT · Random images | runtime | Helped |
| Use Greenkhorn rather than Sinkhorn at small regularization parameter values | ApproxOT · Random images | accuracy | No clear effect |
| Use a KL-divergence analysis to bound marginal error in total variation | Sinkhorn · Sinkhorn projection | iteration complexity: O((\varepsilon')^{-2} \log (s/\ell)) versus O(n\rho(\varepsilon')^{-2} \log (s/\ell)) | Helped |
| Choose each update greedily and perform an exact coordinate line search | Greenkhorn · Sinkhorn projection | runtime | Helped |
Closest papers by meaning
- Screening Sinkhorn Algorithm for Regularized Optimal Transport (NeurIPS 2019 · similarity 0.90)
- On Partial Optimal Transport: Revising the Infeasibility of Sinkhorn and Efficient Gradient Methods (AAAI 2024 · similarity 0.89)
- Computational Optimal Transport: Complexity by Accelerated Gradient Descent Is Better Than by Sinkhorn’s Algorithm (ICML 2018 · similarity 0.89)
- Accelerating Sinkhorn algorithm with sparse Newton iterations (ICLR 2024 · similarity 0.88)
- Safe and Sparse Newton Method for Entropic-Regularized Optimal Transport (NeurIPS 2024 · similarity 0.87)
- Near-Linear Time Generalized Sinkhorn Algorithms for Bounded Genus Graphs (NeurIPS 2026 · similarity 0.87)
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.