{"title":"Sums of distances on graphs and embeddings into Euclidean space","authors":"Stefan Steinerberger","doi":"10.1112/mtk.12198","DOIUrl":null,"url":null,"abstract":"<p>Let <math>\n <semantics>\n <mrow>\n <mi>G</mi>\n <mo>=</mo>\n <mo>(</mo>\n <mi>V</mi>\n <mo>,</mo>\n <mi>E</mi>\n <mo>)</mo>\n </mrow>\n <annotation>$G=(V,E)$</annotation>\n </semantics></math> be a finite, connected graph. We consider a greedy selection of vertices: given a list of vertices <math>\n <semantics>\n <mrow>\n <msub>\n <mi>x</mi>\n <mn>1</mn>\n </msub>\n <mo>,</mo>\n <mi>⋯</mi>\n <mo>,</mo>\n <msub>\n <mi>x</mi>\n <mi>k</mi>\n </msub>\n </mrow>\n <annotation>$x_1, \\dots , x_k$</annotation>\n </semantics></math>, take <math>\n <semantics>\n <msub>\n <mi>x</mi>\n <mrow>\n <mi>k</mi>\n <mo>+</mo>\n <mn>1</mn>\n </mrow>\n </msub>\n <annotation>$x_{k+1}$</annotation>\n </semantics></math> to be any vertex maximizing the sum of distances to the vertices already chosen and iterate, keep adding the “most remote” vertex. The frequency with which the vertices of the graph appear in this sequence converges to a set of probability measures with nice properties. The support of these measures is, generically, given by a rather small number of vertices <math>\n <semantics>\n <mrow>\n <mi>m</mi>\n <mo>≪</mo>\n <mo>|</mo>\n <mi>V</mi>\n <mo>|</mo>\n </mrow>\n <annotation>$m \\ll |V|$</annotation>\n </semantics></math>. We prove that this suggests that the graph <i>G</i> is, in a suitable sense, “<i>m</i>-dimensional” by exhibiting an explicit 1-Lipschitz embedding <math>\n <semantics>\n <mrow>\n <mi>ϕ</mi>\n <mo>:</mo>\n <mi>V</mi>\n <mo>→</mo>\n <msup>\n <mi>ℓ</mi>\n <mn>1</mn>\n </msup>\n <mrow>\n <mo>(</mo>\n <msup>\n <mi>R</mi>\n <mi>m</mi>\n </msup>\n <mo>)</mo>\n </mrow>\n </mrow>\n <annotation>$\\phi : V \\rightarrow \\ell ^1(\\mathbb {R}^m)$</annotation>\n </semantics></math> with good properties.</p>","PeriodicalId":18463,"journal":{"name":"Mathematika","volume":null,"pages":null},"PeriodicalIF":0.8000,"publicationDate":"2023-04-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Mathematika","FirstCategoryId":"100","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1112/mtk.12198","RegionNum":3,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"MATHEMATICS","Score":null,"Total":0}
引用次数: 1
Abstract
Let be a finite, connected graph. We consider a greedy selection of vertices: given a list of vertices , take to be any vertex maximizing the sum of distances to the vertices already chosen and iterate, keep adding the “most remote” vertex. The frequency with which the vertices of the graph appear in this sequence converges to a set of probability measures with nice properties. The support of these measures is, generically, given by a rather small number of vertices . We prove that this suggests that the graph G is, in a suitable sense, “m-dimensional” by exhibiting an explicit 1-Lipschitz embedding with good properties.
期刊介绍:
Mathematika publishes both pure and applied mathematical articles and has done so continuously since its founding by Harold Davenport in the 1950s. The traditional emphasis has been towards the purer side of mathematics but applied mathematics and articles addressing both aspects are equally welcome. The journal is published by the London Mathematical Society, on behalf of its owner University College London, and will continue to publish research papers of the highest mathematical quality.