Anticoncentration for subgraph statistics

From MaRDI portal
Publication:4967964

DOI10.1112/JLMS.12192zbMATH Open1415.05089arXiv1807.05202OpenAlexW2883840497WikidataQ128984309 ScholiaQ128984309MaRDI QIDQ4967964FDOQ4967964


Authors:


Publication date: 11 July 2019

Published in: Journal of the London Mathematical Society (Search for Journal in Brave)

Abstract: Consider integers k,ell such that . Given a large graph G, what is the fraction of k-vertex subsets of G which span exactly ell edges? When G is empty or complete, and ell is zero or , this fraction can be exactly 1. On the other hand, if ell is far from these extreme values, one might expect that this fraction is substantially smaller than 1. This was recently proved by Alon, Hefetz, Krivelevich and Tyomkyn who intiated the systematic study of this question and proposed several natural conjectures. Let . Our main result is that for any k and ell, the fraction of k-vertex subsets that span ell edges is at most logOleft(1ight)left(ell/kight)sqrtk/ell, which is best-possible up to the logarithmic factor. This improves on multiple results of Alon, Hefetz, Krivelevich and Tyomkyn, and resolves one of their conjectures. In addition, we also make some first steps towards some analogous questions for hypergraphs. Our proofs involve some Ramsey-type arguments, and a number of different probabilistic tools, such as polynomial anticoncentration inequalities, hypercontractivity, and a coupling trick for random variables defined on a "slice" of the Boolean hypercube.


Full work available at URL: https://arxiv.org/abs/1807.05202




Recommendations




Cites Work


Cited In (16)





This page was built for publication: Anticoncentration for subgraph statistics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4967964)