arXiv subjects

Jacques Verstraëte

Publications and source records attributed to Jacques Verstraëte.

29 records · Page 2Linked to original sources

arXiv2019

Extremal problems for convex geometric hypergraphs and ordered hypergraphs

math.CO↗

arXiv2019

Hypergraphs not containing a tight tree with a bounded trunk ~II: 3-trees with a trunk of size 2

A tight $r$-tree $T$ is an $r$-uniform hypergraph that has an edge-ordering $e_1, e_2, \dots, e_t$ such that for each $i\geq 2$, $e_i$ has a vertex $v_i$ that does not belong to any previous edge and $e_i-v_i$ is contained in $e_j$ for some $j<i$. Kalai conjectured in 1984 that every $n$-vertex $r$-uniform hypergraph with more than $\frac{t-1}{r}\binom{n}{r-1}$ edges contains every tight $r$-tree $T$ with $t$ edges. A trunk $T'$ of a tight $r$-tree $T$ is a tight subtree $T'$ of $T$ such that vertices in $V(T)\setminus V(T')$ are leaves in $T$. Kalai's Conjecture was proved in 1987 for tight $r$-trees that have a trunk of size one. In a previous paper we proved an asymptotic version of Kalai's Conjecture for all tight $r$-trees that have a trunk of bounded size. In this paper we continue that work to establish the exact form of Kalai's Conjecture for all tight $3$-trees with at least $20$ edges that have a trunk of size two.

math.CO↗

arXiv2019

Ordered and convex geometric trees with linear extremal function

The extremal functions $ex_{\rightarrow}(n,F)$ and $ex_{\cir}(n,F)$ for ordered and convex geometric acyclic graphs $F$ have been extensively investigated by a number of researchers. Basic questions are to determine when $ex_{\rightarrow}(n,F)$ and $ex_{\cir}(n,F)$ are linear in $n$, the latter posed by Braß-Károlyi-Valtr in 2003. In this paper, we answer both these questions for every tree $F$. We give a forbidden subgraph characterization for a family $\cal T$ of ordered trees with $k$ edges, and show that $ex_{\rightarrow}(n,T) = (k - 1)n - {k \choose 2}$ for all $n \geq k + 1$ when $T \in {\cal T}$ and $ex_{\rightarrow}(n,T) = Ω(n\log n)$ for $T \not\in {\cal T}$. We also describe the family of the convex geometric trees with linear Tur\' an number and show that for every convex geometric tree $F$ not in this family, $ex_{\cir}(n,F)= Ω(n\log \log n)$.

math.CO↗

arXiv2018

Extremal problems on ordered and convex geometric hypergraphs

An ordered hypergraph is a hypergraph whose vertex set is linearly ordered, and a convex geometric hypergraph is a hypergraph whose vertex set is cyclically ordered. Extremal problems for ordered and convex geometric graphs have a rich history with applications to a variety of problems in combinatorial geometry. In this paper, we consider analogous extremal problems for uniform hypergraphs, and discover a general partitioning phenomenon which allows us to determine the order of magnitude of the extremal function for various ordered and convex geometric hypergraphs. A special case is the ordered $n$-vertex $r$-graph $F$ consisting of two disjoint sets $e$ and $f$ whose vertices alternate in the ordering. We show that for all $n \geq 2r + 1$, the maximum number of edges in an ordered $n$-vertex $r$-graph not containing $F$ is exactly \[ {n \choose r} - {n - r \choose r}.\] This could be considered as an ordered version of the Erdős-Ko-Rado Theorem, and generalizes earlier results of Capoyleas and Pach and Aronov-Dujmovič-Morin-Ooms-da Silveira.

math.CO↗

arXiv2018

Tight paths in convex geometric hypergraphs

In this paper, we prove a theorem on tight paths in convex geometric hypergraphs, which is asymptotically sharp in infinitely many cases. Our geometric theorem is a common generalization of early results of Hopf and Pannwitz, Sutherland, Kupitz and Perles for convex geometric graphs, as well as the classical Erdős-Gallai Theorem for graphs. As a consequence, we obtain the first substantial improvement on the Turán problem for tight paths in uniform hypergraphs.

math.CO↗

arXiv2017

Hypergraphs not containing a tight tree with a bounded trunk

An $r$-uniform hypergraph is a tight $r$-tree if its edges can be ordered so that every edge $e$ contains a vertex $v$ that does not belong to any preceding edge and the set $e-v$ lies in some preceding edge. A conjecture of Kalai [Kalai], generalizing the Erdős-Sós Conjecture for trees, asserts that if $T$ is a tight $r$-tree with $t$ edges and $G$ is an $n$-vertex $r$-uniform hypergraph containing no copy of $T$ then $G$ has at most $\frac{t-1}{r}\binom{n}{r-1}$ edges. A trunk $T'$ of a tight $r$-tree $T$ is a tight subtree such that every edge of $T-T'$ has $r-1$ vertices in some edge of $T'$ and a vertex outside $T'$. For $r\ge 3$, the only nontrivial family of tight $r$-trees for which this conjecture has been proved is the family of $r$-trees with trunk size one in [FF] from 1987. Our main result is an asymptotic version of Kalai's conjecture for all tight trees $T$ of bounded trunk size. This follows from our upper bound on the size of a $T$-free $r$-uniform hypergraph $G$ in terms of the size of its shadow. We also give a short proof of Kalai's conjecture for tight $r$-trees with at most four edges. In particular, for $3$-uniform hypergraphs, our result on the tight path of length $4$ implies the intersection shadow theorem of Katona [Katona].

math.CO↗

arXiv2017

Stability in the Erdős--Gallai Theorem on cycles and paths, II

The Erdős--Gallai Theorem states that for $k \geq 3$, any $n$-vertex graph with no cycle of length at least $k$ has at most $\frac{1}{2}(k-1)(n-1)$ edges. A stronger version of the Erdős--Gallai Theorem was given by Kopylov: If $G$ is a 2-connected $n$-vertex graph with no cycle of length at least $k$, then $e(G) \leq \max\{h(n,k,2),h(n,k,\lfloor \frac{k-1}{2}\rfloor)\}$, where $h(n,k,a) := {k - a \choose 2} + a(n - k + a)$. Furthermore, Kopylov presented the two possible extremal graphs, one with $h(n,k,2)$ edges and one with $h(n,k,\lfloor \frac{k-1}{2}\rfloor)$ edges. In this paper, we complete a stability theorem which strengthens Kopylov's result. In particular, we show that for $k \geq 3$ odd and all $n \geq k$, every $n$-vertex $2$-connected graph $G$ with no cycle of length at least $k$ is a subgraph of one of the two extremal graphs or $e(G) \leq \max\{h(n,k,3),h(n,k,\frac{k-3}{2})\}$. The upper bound for $e(G)$ here is tight.

math.CO↗

arXiv2016

Full subgraphs

Let $G=(V,E)$ be a graph of density $p$ on $n$ vertices. Following Erdős, Łuczak and Spencer, an $m$-vertex subgraph $H$ of $G$ is called {\em full} if $H$ has minimum degree at least $p(m - 1)$. Let $f(G)$ denote the order of a largest full subgraph of $G$. If $p\binom{n}{2}$ is a non-negative integer, define \[ f(n,p) = \min\{f(G) : \vert V(G)\vert = n, \ \vert E(G)\vert = p\binom{n}{2} \}.\] Erdős, Łuczak and Spencer proved that for $n \geq 2$, \[ (2n)^{\frac{1}{2}} - 2 \leq f(n, {\frac{1}{2}}) \leq 4n^{\frac{2}{3}}(\log n)^{\frac{1}{3}}.\] In this paper, we prove the following lower bound: for $n^{-\frac{2}{3}} <p_n <1-n^{-\frac{1}{7}}$, \[ f(n,p) \geq \frac{1}{4}(1-p)^{\frac{2}{3}}n^{\frac{2}{3}} -1.\] Furthermore we show that this is tight up to a multiplicative constant factor for infinitely many $p$ near the elements of $\{\frac{1}{2},\frac{2}{3},\frac{3}{4},\dots\}$. In contrast, we show that for any $n$-vertex graph $G$, either $G$ or $G^c$ contains a full subgraph on $Ω(\frac{n}{\log n})$ vertices. Finally, we discuss full subgraphs of random and pseudo-random graphs, and several open problems.

math.CO↗

arXiv2016

Stability in the Erdos--Gallai Theorem on cycles and paths

The Erdős-Gallai Theorem states that for $k \geq 2$, every graph of average degree more than $k - 2$ contains a $k$-vertex path. This result is a consequence of a stronger result of Kopylov: if $k$ is odd, $k=2t+1\geq 5$, $n \geq (5t-3)/2$, and $G$ is an $n$-vertex $2$-connected graph with at least $h(n,k,t) := {k-t \choose 2} + t(n -k+ t)$ edges, then $G$ contains a cycle of length at least $k$ unless $G = H_{n,k,t} := K_n - E(K_{n - t})$. In this paper we prove a stability version of the Erdős-Gallai Theorem: we show that for all $n \geq 3t > 3$, and $k \in \{2t+1,2t + 2\}$, every $n$-vertex 2-connected graph $G$ with $e(G) > h(n,k,t-1)$ either contains a cycle of length at least $k$ or contains a set of $t$ vertices whose removal gives a star forest. In particular, if $k = 2t + 1 \neq 7$, we show $G \subseteq H_{n,k,t}$. The lower bound $e(G) > h(n,k,t-1)$ in these results is tight and is smaller than Kopylov's bound $h(n,k,t)$ by a term of $n-t-O(1)$.

math.CO↗

arXiv2013

Turan Problems and Shadows I: Paths and Cycles

A $k$-path is a hypergraph P_k = e_1,e_2,...,e_k such that |e_i \cap e_j| = 1 if |j - i| = 1 and e_i \cap e_j is empty otherwise. A k-cycle is a hypergraph C_k = e_1,e_2,.. ,e_k obtained from a (k-1)-path e_1,e_2,...,e_{k-1} by adding an edge e_k that shares one vertex with e_1, another vertex with e_{k-1} and is disjoint from the other edges. Let ex_r(n,G) be the maximum number of edges in an r-graph with n vertices not containing a given r-graph G. We determine ex_r(n, P_k) and ex_r(n, C_k) exactly for all k \ge 4 and r \ge 3 and $n$ sufficiently large and also characterize the extremal examples. The case k = 3 was settled by Frankl and Füredi. This work is the next step in a long line of research beginning with conjectures of Erd\H os and Sós from the early 1970's. In particular, we extend the work (and settle a recent conjecture) of Füredi, Jiang and Seiver who solved this problem for P_k when r \ge 4 and of Füredi and Jiang who solved it for C_k when r \ge 5. They used the delta system method, while we use a novel approach which involves random sampling from the shadow of an r-graph.

math.CO↗

arXiv2010

The de Bruijn-Erdos Theorem for Hypergraphs

Fix integers $n \ge r \ge 2$. A clique partition of ${[n] \choose r}$ is a collection of proper subsets $A_1, A_2, \ldots, A_t \subset [n]$ such that $\bigcup_i{A_i \choose r}$ is a partition of ${[n] \choose r}$. Let $\cp(n,r)$ denote the minimum size of a clique partition of ${[n] \choose r}$. A classical theorem of de Bruijn and Erd\H os states that $\cp(n, 2) = n$. In this paper we study $\cp(n,r)$, and show in general that for each fixed $r \geq 3$, \[ \cp(n,r) \geq (1 + o(1))n^{r/2} \quad \quad \mbox{as}n \rightarrow \infty.\] We conjecture $\cp(n,r) = (1 + o(1))n^{r/2}$. This conjecture has already been verified (in a very strong sense) for $r = 3$ by Hartman-Mullin-Stinson. We give further evidence of this conjecture by constructing, for each $r \ge 4$, a family of $(1+o(1))n^{r/2}$ subsets of $[n]$ with the following property: no two $r$-sets of $[n]$ are covered more than once and all but $o(n^r)$ of the $r$-sets of $[n]$ are covered. We also give an absolute lower bound $\cp(n,r) \geq {n \choose r}/{q + r - 1 \choose r}$ when $n = q^2 + q + r - 1$, and for each $r$ characterize the finitely many configurations achieving equality with the lower bound. Finally we note the connection of $\cp(n,r)$ to extremal graph theory, and determine some new asymptotically sharp bounds for the Zarankiewicz problem.

math.CO↗