Computer Science > Data Structures and Algorithms
[Submitted on 22 Mar 2024 (v1), last revised 6 Jul 2024 (this version, v2)]
Title:Universal Harmonic Sketching over Turnstile Streams
View PDF HTML (experimental)Abstract:In this work we take a new approach to constructing a universal sketch that focuses on a class of \emph{basis functions} $\{f_s(x)=1-\cos(sx)\mid s>0\}$, so that any $f$-moment can be estimated if $f$ can be expressed as a linear combination of basis functions. We construct and analyze the $\mathsf{SymmetricPoissonTower}$ sketch, which occupies $O(\epsilon^{-2}\log^2(nM\epsilon^{-1}))$ bits and is $\mathcal{F}$-universal for the function class $$\mathcal{F}= \left\{f(x)=cx^2+\int_0^\infty (1-\cos (xs))\,\nu(ds) \mid c\geq 0, \text{$\nu$ is a positive measure}\right\},$$ i.e., given any $f\in \mathcal{F}$, the new sketch $(1\pm\epsilon)$-estimates the $f$-moment with probability 2/3. The family $\mathcal{F}$ includes all the classic frequency moments ($f(z)=|z|^p$, $p\in [0,2]$) as well as a large family of nearly-periodic functions that cannot be estimated with $L_2$-heavy hitter machinery. This new approach to universality requires significantly less space in comparison to previous universal schemes and sheds new light on the full characterization of the class $\mathcal{T}$ of tractable functions.
Submission history
From: Dingyu Wang [view email][v1] Fri, 22 Mar 2024 17:36:19 UTC (727 KB)
[v2] Sat, 6 Jul 2024 03:54:52 UTC (61 KB)
Current browse context:
cs
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.