Research map / Learning theory and algorithms / Sample complexity
WOR and $p$'s: Sketches for $\ell_p$-Sampling Without Replacement
Edith Cohen, Rasmus Pagh, David Woodruff
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 |
|---|---|---|---|
| 2-pass WORp instead of perfect WOR sampling | WORp · l2, Zipf[2], statistic nu^3 | NRMSE: 2.09e-11 -> 2.08e-11 | No clear effect |
| 2-pass WORp instead of perfect WOR sampling | WORp · l1, Zipf[1], statistic nu^3 | NRMSE: 5.73e-03 -> 5.72e-03 | No clear effect |
| WOR (2-pass WORp) instead of perfect WR sampling | WORp · l2, Zipf[2], statistic nu^3 | NRMSE: 1.16e-04 -> 2.08e-11 | Improved |
| WOR (2-pass WORp) instead of perfect WR sampling | WORp · l1, Zipf[1], statistic nu^3 | NRMSE: 3.59e-01 -> 5.72e-03 | Improved |
| WOR (2-pass WORp) instead of perfect WR sampling | WORp · l1, Zipf[2], statistic nu | NRMSE: 9.51e-03 -> 1.60e-03 | Improved |
| WOR (2-pass WORp) instead of perfect WR sampling | WORp · l2, Zipf[2], statistic nu^2 | NRMSE: 7.96e-05 -> 1.25e-07 | Improved |
| 1-pass WORp (estimated weights) instead of 2-pass WORp | WORp · l2, Zipf[2], statistic nu^2 | NRMSE: 1.25e-07 -> 1.14e-02 | Hurt |
| 1-pass WORp instead of 2-pass WORp | WORp · l1, Zipf[2], statistic nu | NRMSE: 1.60e-03 -> 2.79e-02 | Hurt |
| 1-pass WORp instead of 2-pass WORp | WORp · l1, Zipf[1], statistic nu^3 | NRMSE: 5.72e-03 -> 5.14e-03 | No clear effect |
| 1-pass WORp instead of perfect WR | WORp · l1, Zipf[2], statistic nu^3 | NRMSE: 3.45e-04 -> 5.11e-05 | Improved |
| 1-pass WORp instead of perfect WR | WORp · l2, Zipf[2], statistic nu^3 | NRMSE: 1.16e-04 -> 1.06e-03 | Hurt |
Closest papers by meaning
- Composable Sketches for Functions of Frequencies: Beyond the Worst Case (ICML 2020 · similarity 0.88)
- Sampling Sketches for Concave Sublinear Functions of Frequencies (NeurIPS 2019 · similarity 0.86)
- A Separation in Heavy-Tailed Sampling: Gaussian vs. Stable Oracles for Proximal Samplers (NeurIPS 2024 · similarity 0.85)
- Turnstile $\ell_p$ leverage score sampling with applications (ICML 2024 · similarity 0.84)
- Fast Sampling-Based Sketches for Tensors (ICML 2024 · similarity 0.83)
- Optimal bounds for $\ell_p$ sensitivity sampling via $\ell_2$ augmentation (ICML 2024 · similarity 0.82)
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.
