Home /  PSDS Seminar: Probabilistic methods for stable sets in hypergraphs

Seminar

PSDS Seminar: Probabilistic methods for stable sets in hypergraphs March 06, 2025 (11:00 AM PST - 12:00 PM PST)
Parent Program:
Location: SLMath: Online/Virtual, Baker Board Room
Speaker(s) Jacques Verstraete (University of California, San Diego)
Description No Description
Keywords and Mathematics Subject Classification (MSC)
Primary Mathematics Subject Classification No Primary AMS MSC
Secondary Mathematics Subject Classification No Secondary AMS MSC
Video
No Video Uploaded
Abstract/Media

Zoom Link

Let $\mathcal{S} = \{S_1,S_2,\dots,S_n\}$ be a family of finite sets. A maximal independent set in $\mathcal{S}$ is a set $X \subseteq \bigcup_{i = 1}^n S_i$ such that for all $i = 1,2,\dots,n$, $S_i \not \subseteq X$. We consider a variety of methods for finding maximal independent sets in graphs using randomness, and in particular, a simple random process for producing a maximal independent set in a family $\mathcal{S}$ of sets based on restrictions on the intersections between the sets in the family. In particular, the following theorem is a corollary, addressing a long-standing problem of Segre (1952) and improving earier results of Kim and Vu: 

\bigskip

{\bf Theorem.} There exists a constant $c > 0$ such that in any projective plane of order $q$, there exists a set $X$ of at most $c\sqrt{q}\log q$ points such that every line contains at most two of the points, but the addition of any point to $X$ gives a line containing two elements of $X$ plus the additional point. 

\bigskip

In the language of finite geometry, this is to say that every projective plane of order $q$ contains a {\em maximal arc} of size of order at most $c\sqrt{q}\log q$. 

No Notes/Supplements Uploaded No Video Files Uploaded