Research map / Learning theory and algorithms / Kernels and regression theory
Tighter Expected Generalization Error Bounds via Wasserstein Distance
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 single-letter Wasserstein distances between the hypothesis distribution given one sample and its marginal | Expected generalization error bounds in the standard setting · Standard setting | expected generalization error bound | Helped |
| Use single-letter Wasserstein distances instead of the corresponding relative-entropy or mutual-information bounds | Expected generalization error bounds with bounded loss · Standard setting | expected generalization error bound | Helped |
| Use random-subset Wasserstein distances instead of the corresponding relative-entropy bounds | Expected generalization error bounds with bounded loss · Standard setting | expected generalization error bound | Helped |
| Use single-letter Wasserstein bounds in place of full-dataset Wasserstein bounds | Gaussian location model · Gaussian location model | rate of expected generalization error bound: O(√(σ²d)/n) versus O(√(σ²d/n)) | Helped |
| Use single-letter Wasserstein bounds in place of single-letter mutual-information bounds | Expected generalization error bounds in the randomized-subsample setting · Randomized-subsample setting | expected generalization error bound | Helped |
| Use randomized-subsample random-subset Wasserstein bounds rather than the single-letter bound | Expected generalization error bounds in the randomized-subsample setting · Randomized-subsample setting | expected generalization error bound | Hurt |
| Use the backward channel and Lipschitz geometry of the sample space | Expected generalization error bounds · Standard setting | expected generalization error bound | Helped |
| Use bounded loss and the discrete metric to derive total-variation bounds | Standard-setting relative-entropy generalization bounds · Standard setting | vacuity of expected generalization error bound | Helped |
Closest papers by meaning
- Exactly Tight Information-theoretic Generalization Bounds via Binary Jensen-Shannon Divergence (ICML 2025 · similarity 0.87)
- Statistical and Computational Guarantees of Kernel Max-Sliced Wasserstein Distances (ICML 2025 · similarity 0.86)
- Fantastic Generalization Measures are Nowhere to be Found (ICLR 2024 · similarity 0.86)
- Universal generalization guarantees for Wasserstein distributionally robust models (ICLR 2025 · similarity 0.86)
- A New Robust Partial p-Wasserstein-Based Metric for Comparing Distributions (ICML 2024 · similarity 0.86)
- Scalable Approximation Algorithms for $p$-Wasserstein Distance and Its Variants (ICML 2025 · similarity 0.86)
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.