Research map / Learning theory and algorithms / Optimal transport
On Scalable and Efficient Computation of Large Scale Optimal Transport
Yujia Xie, Minshuo Chen, Haoming Jiang, Tuo Zhao, Hongyuan Zha
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use DASPOT instead of DeepJDOT | domain adaptation · MNIST → USPS | accuracy: 95.7% -> 97.5% | Improved |
| Use DASPOT instead of DeepJDOT | domain adaptation · USPS → MNIST | accuracy: 96.4% -> 96.5% | Improved |
| Use DASPOT instead of DeepJDOT | domain adaptation · SVHN → MNIST | accuracy: 96.7% -> 96.2% | Hurt |
| Use DASPOT instead of DeepJDOT | domain adaptation · MNIST → MNISTM | accuracy: 92.4% -> 94.9% | Improved |
| Use SPOT rather than ROT | Gaussian OT, 2D · Gaussian OT, 2D | convergence to Wasserstein distance: 0.6%, 0.3%, and 0.3% relative errors | Improved |
| Use DASPOT instead of DeepJDOT | MNIST → USPS domain adaptation · MNIST → USPS | running time to 95% accuracy: 518s -> 169s | Improved |
| Increase the number of hidden units per layer, up to 2^{10} (ablation) | SPOT · Gaussian OT, 2D | relative error | No clear effect |
| Increase the entropy regularization coefficient epsilon (ablation) | neural ODE SPOT density recovery · Gaussian and Gaussian-mixture OT | joint-density smoothness | Improved |
| Use the semantic-aware edge cost to encourage pairing (ablation) | SPOT sample generation · MNIST–MNISTM | paired-image quality | Improved |
Closest papers by meaning
- Stochastic Optimization for Large-scale Optimal Transport (NeurIPS 2016 · similarity 0.90)
- Progressive Entropic Optimal Transport Solvers (NeurIPS 2024 · similarity 0.88)
- On Partial Optimal Transport: Revising the Infeasibility of Sinkhorn and Efficient Gradient Methods (AAAI 2024 · similarity 0.88)
- Geometry-Aware Optimal Transport: Fast Intrinsic Dimension and Wasserstein Distance Estimation (NeurIPS 2026 · similarity 0.87)
- Overcoming Spurious Solutions in Semi-Dual Neural Optimal Transport: A Smoothing Approach for Learning the Optimal Transport Plan (ICML 2025 · similarity 0.87)
- Expectile Regularization for Fast and Accurate Training of Neural Optimal Transport (NeurIPS 2024 · 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.
