Research map / Learning theory and algorithms / Optimal transport
Mixed-Curvature Tree-Sliced Wasserstein Distance
Duy-Tung Pham, Viet-Hoang Tran, Thieu Vo, Tan Minh Nguyen
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 |
|---|---|---|---|
| replace product-space or ambient slicing with MCTSW in gradient flow | sliced Wasserstein baselines · Gradient flow (6 WND mixture) | log W2: 0.31 -> -3.65 | Improved |
| use MCTSW as uniformity loss in a mixed-curvature latent space | SSGE · Cora (graph SSL) | accuracy: 79.55 -> 79.86 | Improved |
| use mixed-curvature latent instead of constant-curvature (E/H/S) TSW | TSW-SSGE · Cora (graph SSL) | accuracy: 78.33 -> 79.86 | Improved |
| replace KL regularizer with MCTSW in mixed-curvature VAE | M-VAE · CIFAR-10 VAE | test BCE: 0.6419 -> 0.6000 | Improved |
| MCTSW in mixed-curvature latent vs SW-based constant-curvature variants | HSW-VAE · CIFAR-10 VAE | test BCE: 0.6012 -> 0.6000 | Improved |
| increase max curvature magnitude to 4 (from Euclidean 0) (ablation) | MCTSW-SSGE · Cora (graph SSL) | accuracy: 77.9 -> 79.9 -> 77.3 (maxK 0, 4, 8) | Mixed |
| increase number of subspaces from 8 to 32 (ablation) | MCTSW-SSGE · Cora (graph SSL) | accuracy: 76.7 -> 79.9 | Improved |
| increase number of subspaces from 32 to 128 (ablation) | MCTSW-SSGE · Cora (graph SSL) | accuracy: 79.9 -> 78.7 | Hurt |
| vary number of rays per tree k from 4 to 512 (ablation) | MCTSW-SSGE · Cora (graph SSL) | accuracy: 79.6 (k=4), 79.9 (k=8), 79.5 (k=512) | No clear effect |
| use spherical projection instead of orthogonal projection (with orthogonal split) (ablation) | MCTSW · Gradient flow (6 WND mixture) | log W2: -1.7517 -> -3.6576 | Improved |
| use orthogonal split instead of spherical split (with spherical projection) (ablation) | MCTSW · Gradient flow (6 WND mixture) | log W2: -3.5546 -> -3.6576; run time 1110 -> 303 | Improved |
| use simplified tree sampling (closed-form distance to ray) instead of numerical ternary search (ablation) | MCTSW · Gradient flow (6 WND mixture) | log W2: -3.6576 vs -3.6569; run time 303 vs 3479 | No clear effect |
Closest papers by meaning
- Tree-Sliced Wasserstein Distance with Nonlinear Projection (ICML 2025 · similarity 0.91)
- Distance-Based Tree-Sliced Wasserstein Distance (ICLR 2025 · similarity 0.90)
- Spherical Tree-Sliced Wasserstein Distance (ICLR 2025 · similarity 0.88)
- Tree-Sliced Wasserstein Distance: A Geometric Perspective (ICML 2025 · similarity 0.88)
- Stereographic Spherical Sliced Wasserstein Distances (ICML 2024 · similarity 0.88)
- Mixed-curvature decision trees and random forests (ICML 2025 · 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.
