Research map / Learning theory and algorithms / Optimal transport
Mapping Estimation for Discrete Optimal Transport
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 |
|---|---|---|---|
| jointly learn coupling and linear map (OTLin) instead of classic OT | OT domain adaptation · Office-Caltech mean | accuracy (%): 68.6 -> 84.1 | Helped |
| jointly learn coupling and linear map (OTLin) instead of entropic OTE | OT domain adaptation · Office-Caltech mean | accuracy (%): 74.6 -> 84.1 | Helped |
| linear map instead of kernel map (ablation) | OTLin vs OTKer · Office-Caltech mean | accuracy (%): 79.6 -> 84.1 | Helped |
| linear map instead of kernel map (ablation) | OTLin vs OTKer · Moons 90 degrees | accuracy (%): 56.3 -> 67.9 | Helped |
| add bias term to map (OTLinB vs OTLin) (ablation) | OTLin · Office-Caltech mean | accuracy (%): 84.1 -> 84.1 | No clear effect |
| add bias term to kernel map (OTKerB vs OTKer) (ablation) | OTKer · Office-Caltech mean | accuracy (%): 79.6 -> 80.0 | Helped |
| evaluate learned map T instead of coupling gamma (ablation) | OTLin · Office-Caltech mean | accuracy (%): 84.1 -> 84.1 | No clear effect |
| proposed joint learning versus classic OT at large rotation | OTLin · Moons 90 degrees | accuracy (%): 51.0 -> 67.9 | Helped |
| proposed OTLin versus unadapted 1NN | OTLin · Moons 90 degrees | accuracy (%): 19.4 -> 67.9 | Helped |
| OT-based methods at small rotation | OT variants · Moons 10-40 degrees | accuracy (%): 100.0 | No clear effect |
| proposed OTLin versus OTE/L1L2 on harder rotation | OTLin · Moons 70 degrees | accuracy (%): 89.0 -> 88.0 | Mixed |
| kernel map vs linear map on individual Office-Caltech tasks (ablation) | OTKer · A->C | accuracy (%): 81.8 -> 84.4 | Helped |
Closest papers by meaning
- Progressive Entropic Optimal Transport Solvers (NeurIPS 2024 · similarity 0.89)
- Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan (ICML 2025 · similarity 0.88)
- Low-Rank Optimal Transport through Factor Relaxation with Latent Coupling (NeurIPS 2024 · similarity 0.88)
- On the Private Estimation of Smooth Transport Maps (ICML 2025 · similarity 0.88)
- Hierarchical Refinement: Optimal Transport to Infinity and Beyond (ICML 2025 · similarity 0.88)
- Stochastic Optimization in Semi-Discrete Optimal Transport: Convergence Analysis and Minimax Rate (NeurIPS 2025 · similarity 0.88)
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.