Computer Science > Computational Geometry
[Submitted on 30 Jul 2013]
Title:A Size-Sensitive Discrepancy Bound for Set Systems of Bounded Primal Shatter Dimension
View PDFAbstract:Let $(X,§)$ be a set system on an $n$-point set $X$. The \emph{discrepancy} of $§$ is defined as the minimum of the largest deviation from an even split, over all subsets of $S \in §$ and two-colorings $\chi$ on $X$. We consider the scenario where, for any subset $X' \subseteq X$ of size $m \le n$ and for any parameter $1 \le k \le m$, the number of restrictions of the sets of $§$ to $X'$ of size at most $k$ is only $O(m^{d_1} k^{d-d_1})$, for fixed integers $d > 0$ and $1 \le d_1 \le d$ (this generalizes the standard notion of \emph{bounded primal shatter dimension} when $d_1 = d$). In this case we show that there exists a coloring $\chi$ with discrepancy bound $O^{*}(|S|^{1/2 - d_1/(2d)} n^{(d_1 - 1)/(2d)})$, for each $S \in §$, where $O^{*}(\cdot)$ hides a polylogarithmic factor in $n$. This bound is tight up to a polylogarithmic factor \cite{Mat-95, Mat-99} and the corresponding coloring $\chi$ can be computed in expected polynomial time using the very recent machinery of Lovett and Meka for constructive discrepancy minimization \cite{LM-12}. Our bound improves and generalizes the bounds obtained from the machinery of Har-Peled and Sharir \cite{HS-11} (and the follow-up work in \cite{SZ-12}) for points and halfspaces in $d$-space for $d \ge 3$.
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.