Sharper Bounds for $\ell_p$ Sensitivity Sampling

Authors: David Woodruff, Taisuke Yasuda

ICML 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical In this work, we show the first bounds for sensitivity sampling for ℓ𝑝subspace embeddings for 𝑝 = 2 that improve over the general S𝑑bound, achieving a bound of roughly S2/𝑝for 1 𝑝< 2 and S2 2/𝑝for 2 < 𝑝< . For 1 𝑝< 2, we show that this bound is tight, in the sense that there exist matrices for which S2/𝑝samples is necessary. Furthermore, our techniques yield further new results in the study of sampling algorithms, showing that the root leverage score sampling algorithm achieves a bound of roughly 𝑑for 1 𝑝< 2, and that a combination of leverage score and sensitivity sampling achieves an improved bound of roughly 𝑑2/𝑝S2 4/𝑝for 2 < 𝑝< . Our sensitivity sampling results yield the best known sample complexity for a wide class of structured matrices that have small ℓ𝑝sensitivity. ... Our work introduces a new analysis for sensitivity sampling for ℓ𝑝subspace embeddings, which breaks a previous general sampling barrier of 𝑂(πœ€ 2S𝑝(A)𝑑) samples via a simple union bound argument, to obtain an improved bound of 𝑂(πœ€ 2S𝑝(A)2/𝑝) samples for 𝑝< 2 and 𝑂(πœ€ 2S𝑝(A)2 2/𝑝) samples for 𝑝> 2.
Researcher Affiliation Academia 1Computer Science Department, Carnegie Mellon University, Pittsburgh, PA, US. Correspondence to: Taisuke Yasuda <taisukey@cs.cmu.edu>.
Pseudocode No The paper describes algorithmic procedures in prose but does not include any explicitly labeled 'Pseudocode' or 'Algorithm' blocks.
Open Source Code No The paper does not contain any statement about releasing source code for the described methodology, nor does it provide any links to a code repository.
Open Datasets No The paper is theoretical and does not conduct experiments on a specific dataset, therefore no public dataset access information is provided.
Dataset Splits No The paper is theoretical and does not describe empirical experiments or data processing, thus no training/test/validation dataset splits are provided.
Hardware Specification No The paper is theoretical and does not describe empirical experiments, therefore no hardware specifications are mentioned.
Software Dependencies No The paper is theoretical and does not describe empirical experiments, therefore no software dependencies with version numbers are listed.
Experiment Setup No The paper is theoretical and does not describe empirical experiments, therefore no specific experimental setup details or hyperparameters are provided.