Research map / Learning theory and algorithms / Sample complexity
Statistical and Topological Properties of Sliced Probability Divergences
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 |
|---|---|---|---|
| slice the divergence (Sliced-Wasserstein instead of Wasserstein) | Wasserstein · Gaussian samples, varying dimension | sample-complexity convergence rate | Helped |
| slice the Sinkhorn divergence (Sliced-Sinkhorn) | Sinkhorn divergence · Gaussian samples, varying d and epsilon | sample-complexity convergence rate | Helped |
| slice the Sinkhorn divergence (Sliced-Sinkhorn) | Sinkhorn divergence · MNIST/CIFAR-10 two-sample | execution time: almost 130 times faster | Helped |
| slice the Sinkhorn divergence (Sliced-Sinkhorn) | Sinkhorn divergence · Gaussian samples, increasing dimension | Sinkhorn iterations until convergence | Helped |
| use sliced divergences in two-sample testing on real data | Wasserstein and Sinkhorn · MNIST and CIFAR-10 | samples needed to converge to zero | Helped |
| increase number of Monte Carlo projections L (ablation) | Sliced-Wasserstein · Gaussian samples, d=2 to 100 | Monte Carlo error | Helped |
Closest papers by meaning
- Computing Divergences between Discrete Decomposable Models (AAAI 2023 · similarity 0.86)
- On Slicing Optimality for Mutual Information (NeurIPS 2023 · similarity 0.86)
- Approximating f -Divergences with Rank Statistics (ICML 2026 · similarity 0.86)
- Divergence Frontiers for Generative Models: Sample Complexity, Quantization Effects, and Frontier Integrals (NeurIPS 2021 · similarity 0.86)
- Hierarchical Integral Probability Metrics: A distance on random probability measures with low sample complexity (ICML 2024 · similarity 0.84)
- Optimization Can Learn Johnson Lindenstrauss Embeddings (NeurIPS 2024 · similarity 0.84)
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.