Computer Science > Discrete Mathematics
[Submitted on 8 May 2021 (v1), revised 13 Jul 2021 (this version, v2), latest version 29 Nov 2021 (v3)]
Title:On set systems definable in sparse graph classes, discrepancy, and quantifier elimination
View PDFAbstract:We study the (hereditary) discrepancy of definable set systems, which is a natural measure for their inherent complexity and approximability. We establish a strong connection between the hereditary discrepancy and quantifier elimination over signatures with only unary relation and function symbols. We prove that set systems definable in theories (over such signatures) with quantifier elimination have constant hereditary discrepancy. We derive that set systems definable in bounded expansion classes, which are very general classes of uniformly sparse graphs, have bounded hereditary discrepancy. We also derive that nowhere dense classes, which are more general than bounded expansion classes, in general do not admit quantifier elimination over a signature extended by an arbitrary number of unary function symbols. We show that the set systems on a ground set $U$ definable on a monotone nowhere dense class of graphs $\mathscr C$ have hereditary discrepancy at most $|U|^{c}$ (for some~$c<1/2$) and that, on the contrary, for every monotone somewhere dense class $\mathscr C$ and for every positive integer $d$ there is a set system of $d$-tuples definable in $\mathscr C$ with discrepancy $\Omega(|U|^{1/2})$.
We further prove that if $\mathscr C$ is a class of graphs with bounded expansion and $\phi(\bar x;\bar y)$ is a first-order formula, then we can compute in polynomial time, for an input graph $G\in\mathscr C$, a mapping $\chi:V(G)^{|\bar x|}\rightarrow\{-1,1\}$ witnessing the boundedness of the discrepancy of the set-system defined by~$\phi$, an $\varepsilon$-net of size $\mathcal{O}(1/\varepsilon)$, and an $\varepsilon$-approximation of size $\mathcal{O}(1/\varepsilon)$.
Submission history
From: Sebastian Siebertz [view email][v1] Sat, 8 May 2021 13:05:31 UTC (247 KB)
[v2] Tue, 13 Jul 2021 08:19:28 UTC (429 KB)
[v3] Mon, 29 Nov 2021 08:50:49 UTC (275 KB)
Current browse context:
cs.DM
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.