Search

How small can a set of vertices in the $n$-dimensional hypercube $Q_n$ be if it meets every copy of $Q_d$? The asymptotic density of such a set (for $d$ fixed and $n$ large) is denoted by $\gamma_d$. It is easy to see that $\gamma_d \leq 1/(d+1)$, and it is known that $\gamma_d=1/(d+1)$ for $d \leq 2$, but it was recently shown that $\gamma_d < 1/(d+1)$ for $d \geq 8$. In this paper we show that the latter phenomenon also holds for $d=7$ and $d=6$., Comment: 8 pages

Abstract Background: Pulmonary function tests (PFTs) and computed tomography (CT) imaging are vital in diagnosing, managing, and monitoring lung diseases. A common issue in practice is the lack of access to recorded pulmonary functions despite available chest CT scans. Purpose: To develop and validate a deep learning algorithm for predicting pulmonary function directly from chest CT scans. Methods: The development cohort came from the Pittsburgh Lung Screening Study (PLuSS) (n=3619). The validation cohort came from the Specialized Centers of Clinically Oriented Research (SCCOR) in COPD (n=662). A deep learning model called BeyondCT, combining a three-dimensional (3D) convolutional neural network (CNN) and Vision Transformer (ViT) architecture, was used to predict forced vital capacity (FVC) and forced expiratory volume in one second (FEV1) from non-contrasted inspiratory chest CT scans. A 3D CNN model without ViT was used for comparison. Subject demographics (age, gender, smoking status) were also incorporated into the model. Performance was compared to actual PFTs using mean absolute error (MAE, L), percentage error, and R square. Results: The 3D-CNN model achieved MAEs of 0.395 L and 0.383 L, percentage errors of 13.84% and 18.85%, and R square of 0.665 and 0.679 for FVC and FEV1, respectively. The BeyondCT model without demographics had MAEs of 0.362 L and 0.371 L, percentage errors of 10.89% and 14.96%, and R square of 0.719 and 0.727, respectively. Including demographics improved performance (p<0.05), with MAEs of 0.356 L and 0.353 L, percentage errors of 10.79% and 14.82%, and R square of 0.77 and 0.739 for FVC and FEV1 in the test set. Conclusion: The BeyondCT model showed robust performance in predicting lung function from non-contrast inspiratory chest CT scans., Comment: 5 tables, 7 figures,22 pages

The `odd cover number' of a complete graph is the smallest size of a family of complete bipartite graphs that covers each edge an odd number of times. For $n$ odd, Buchanan, Clifton, Culver, Nie, O'Neill, Rombach and Yin showed that the odd cover number of $K_n$ is equal to $(n+1)/2$ or $(n+3)/2$, and they conjectured that it is always $(n+1)/2$. We prove this conjecture. For $n$ even, Babai and Frankl showed that the odd cover number of $K_n$ is always at least $n/2$, and the above authors and Radhakrishnan, Sen and Vishwanathan gave some values of $n$ for which equality holds. We give some new examples. Our constructions arise from some very symmetric constructions for the corresponding problem for complete hypergraphs. Thus the odd cover number of the complete 3-graph $K_n^{(3)}$ is the smallest number of complete 3-partite 3-graphs such that each 3-set is in an odd number of them. We show that the odd cover number of $K_n^{(3)}$ is exactly $n/2$ for even $n$, and we show that for odd $n$ it is $(n-1)/2$ for infinitely many values of $n$. We also show that for $r=3$ and $r=4$ the odd cover number of $K_n^{(r)}$ is strictly less than the partition number, answering a question of Buchanan, Clifton, Culver, Nie, O'Neill, Rombach and Yin for those values of $r$., Comment: 7 pages

Fern\'andez-Bret\'on, Sarmiento and Vera showed that whenever a direct sum of sufficiently many copies of ${\mathbb Z}_4$, the cyclic group of order 4, is countably coloured there are arbitrarily large finite sets $X$ whose sumsets $X+X$ are monochromatic. They asked if the elements of order 4 are necessary, in the following strong sense: if $G$ is an abelian group having no elements of order 4, is it always the case there there is a countable colouring of $G$ for which there is not even a monochromatic sumset $X+X$ with $X$ of size 2? Our aim in this short note is to show that this is indeed the case.

A set $X$ is called Euclidean Ramsey if, for any $k$ and sufficiently large $n$, every $k$-colouring of $\mathbb{R}^n$ contains a monochromatic congruent copy of $X$. This notion was introduced by Erd\H{o}s, Graham, Montgomery, Rothschild, Spencer and Straus. They asked if a set is Ramsey if and only if it is spherical, meaning that it lies on the surface of a sphere. It is not too difficult to show that if a set is not spherical then it is not Euclidean Ramsey either, but the converse is very much open despite extensive research over the years. On the other hand, the block sets conjecture is a purely combinatorial, Hales-Jewett type of statement, concerning `blocks in large products', introduced by Leader, Russell and Walters. If true, the block sets conjecture would imply that every transitive set (a set whose symmetry group acts transitively) is Euclidean Ramsey. As for the question above, the block sets conjecture remains very elusive, being known only in a few cases. In this paper we show that the sizes of the blocks in the block sets conjecture cannot be bounded, even for templates over the alphabet of size $3$. We also show that for the first non-trivial template, namely $123$, the blocks may be taken to be of size $2$ (for any number of colours). This is best possible; all previous bounds were `tower-type' large., Comment: 10 pages

A subgraph of the $n$-dimensional hypercube is called 'layered' if it is a subgraph of a layer of some hypercube. In this paper we show that there exist subgraphs of the cube of arbitrarily large girth that are not layered. This answers a question of Axenovich, Martin and Winter. Perhaps surprisingly, these subgraphs may even be taken to be induced., Comment: 11 pages

An $r$-daisy is an $r$-uniform hypergraph consisting of the six $r$-sets formed by taking the union of an $(r-2)$-set with each of the 2-sets of a disjoint 4-set. Bollob\'as, Leader and Malvenuto, and also Bukh, conjectured that the Tur\'an density of the $r$-daisy tends to zero as $r \to \infty$. In this paper we disprove this conjecture. Adapting our construction, we are also able to disprove a folklore conjecture about Tur\'an densities of hypercubes. For fixed $d$ and large $n$, we show that the smallest set of vertices of the $n$-dimensional hypercube $Q_n$ that meets every copy of $Q_d$ has asymptotic density strictly below $1/(d+1)$, for all $d \geq 8$. In fact, we show that this asymptotic density is at most $c^d$, for some constant $c<1$. As a consequence, we obtain similar bounds for the edge-Tur\'an densities of hypercubes. We also answer some related questions of Johnson and Talbot, and disprove a conjecture made by Bukh and by Griggs and Lu on poset densities., Comment: 14 pages. Minor corrections made since last version

How many graphs on an $n$-point set can we find such that any two have connected intersection? Berger, Berkowitz, Devlin, Doppelt, Durham, Murthy and Vemuri showed that the maximum is exactly $1/2^{n-1}$ of all graphs. Our aim in this short note is to give a 'directed' version of this result; we show that a family of oriented graphs such that any two have strongly-connected intersection has size at most $1/3^n$ of all oriented graphs. We also show that a family of graphs such that any two have Hamiltonian intersection has size at most $1/2^n$ of all graphs, verifying a conjecture of the above authors., Comment: 5 pages

Suppose that $A$ and $B$ are sets in $\mathbb{R}^d$, and we form the sumset of $A$ with $n$ random points of $B$. Given the volumes of $A$ and $B$, how should we choose them to minimize the expected volume of this sumset? Our aim in this paper is to show that we should take $A$ and $B$ to be Euclidean balls. We also consider the analogous question in the torus $\mathbb{T}^d$, and we show that in this case the optimal choices of $A$ and $B$ are bands, in other words, sets of the form $[x,y]\times\mathbb{T}^{d-1}$. We also give stability versions of our results.