Graph Theory Uncovers the Roots of Perfection

Total Page:16

File Type:pdf, Size:1020Kb

Graph Theory Uncovers the Roots of Perfection N EWS F OCUS 65 were later named “Berge graphs”) would be 64 MATHEMATICS perfect (see figure). But he couldn’t prove it, 63 and his speculation became the SPGC. At 62 the same time, Berge also ventured a less 61 Graph Theory Uncovers the definitive pronouncement known as the 60 “weak” perfect graph conjecture. 59 Seymour, along with co-authors G. Neil 58 Roots of Perfection Robertson of Ohio State University in 57 Columbus and Robin Thomas of the Georgia A newly minted proof tells how to recognize which arrangements of 56 Institute of Technology in Atlanta, began points and lines are the crème de la crème 55 working on the SPGC in 2000. They were 54 motivated in part by a grant from the private- 53 To some, perfection is priceless. But for phone network based on a perfect graph ly funded American Institute of Mathematics 52 four graph theorists, it has a very specific could run with optimum efficiency even if in Palo Alto, California, which promotes 51 value. If their solution to one of the oldest some of its transmitters were knocked out work on high-profile unsolved problems. Al- 50 problems in their discipline—a classifica- (although it would cover less area). though mathematicians have identified a 49 tion of so-called perfect graphs—holds up, A perfect graph is like a perfect chocolate dizzying variety of perfect graphs—96 types 48 they will reap a $10,000 bounty. cake: It might be easy to describe, but it’s hard at last count—the three researchers focused 47 The strong perfect graph conjecture to produce a recipe. In 1960, however, Claude on just two types, called bipartite graphs and 46 (SPGC) has perplexed mathematicians for Berge, a mathematician at the Centre National line graphs, as well as their antigraphs. Fol- 45 more than 40 years. “It’s a problem that ev- de la Recherche Scientifique in Paris, did just lowing a strategy suggested by another per- 44 eryone in graph theory knows about, and that. He noticed that every imperfect graph he fect-graph aficionado, Gerard Cornuejols of 43 some people in related areas, particularly could find contained either an “odd hole” or Carnegie Mellon University in Pittsburgh, 42 linear programming,” says Paul Seymour of an “odd anti-hole.” An odd hole is a ring of an Pennsylvania, they proved that any Berge 41 Princeton University, who announced the odd number (at least 5) of nodes, each linked graph that is not one of these types can be 40 proof at a meeting of the Canadian Mathe- to its two neighbors but not to any other node decomposed into smaller pieces that are. 39 matical Society last month. Its solution in the ring. An anti-hole is the reverse: Each Cornuejols had done more than just sug- 38 might enable mathematicians to quickly node is connected to every other node in the gest a strategy. He put his money where his 37 identify perfect graphs, which have proper- ring except its neighbors. mouth is, offering $5000 for the proof of a 36 ties that make otherwise intractable prob- Berge boldly conjectured that any graph particular step that had eluded him, as well 35 lems involving networks easy to solve. that avoided these two flaws (such graphs as $5000 for completing the proof of the full 34 The graphs in question consist of noth- SPGC. To collect the prize, Seymour and his 33 ing more than dots and lines. Each line con- collaborators will have to get their work pub- 32 nects exactly two dots, or nodes. The SPGC lished, which might take a year or more 31 grew out of mathematicians’ fascination channels clique while other graph theorists scrutinize a proof 30 with coloring graphs in such a way that no that will likely run to 150 or 200 pages. 29 two nodes of the same color are connected, The early betting is that they will collect 28 a problem rooted in the real-world business the prize. “I don’t know the details of the 27 of coloring maps. When Wolfgang Haken proof, but I trust that it is basically correct,” 26 and Kenneth Appel proved the famous Four- says László Lovász of Microsoft Research, 25 Color Theorem for planar maps in 1976, who proved the weak perfect graph theorem 24 they did it by means of graph theory. chi = omega = 3 in 1972. “In the first version of a complicat- 23 Coloring problems make sense for other ed and long proof like this, there are always 22 kinds of graphs as well. In a cell-phone some gaps and, at the same time, often sub- 21 network, for example, the nodes are trans- perfect stantial possibilities for simplification. But I 20 mitters, the lines connect any two transmit- do know the general plan, and I have no 19 ters whose ranges overlap, and the colors 12 doubt that it is now working.” Late last 18 correspond to channels. Coloring the net- month, Seymour and his student Maria 13 3 17 work amounts to assigning channels so that 12 Chudnovsky presented more details at a 16 no adjacent transmitters broadcast on the workshop in Oberwolfach, Germany. András 11 4 15 same channel. Of course, the phone compa- chi = 4 Sebö of the Institut d’Informatique et Mathé- 14 ny would want to use the smallest possible omega = 3 matiques Appliquées in Grenoble, France, 13 number of channels, which is called the 10 5 who co-organized the workshop, says the ex- 12 (χ) chromatic number chi of the network. 9 6 cellent track record of the authors, and the 11 It’s easy to see that any group of nodes 8 7 ease with which Seymour and Chudnovsky 10 that are all connected to one another must all fielded all questions, leaves “not much 9 be different colors. Graph theorists call such imperfect doubt” that the proof will hold up. 8 a dense web of nodes a clique. Thus, in any Before the workshop, Seymour e-mailed 7 graph, chi has to be at least as large as the Net watch. In a web of cell-phone transmit- news of the proof to Berge, who proposed the 6 size of the biggest clique, a number known ters based on a perfect graph, the smallest problem. He later heard that Berge, who is ω 5 as omega ( ). In a perfect graph, in fact, chi number of channels needed to avoid interfer- seriously ill, had the message read to him in 4 and omega are exactly equal. More than that, ence (chi) equals the largest number of inter- the hospital. “He was happy,” Seymour says. 3 they stay equal no matter how many nodes connected nodes (omega). Adding two trans- –DANA MACKENZIE 2 you knock out of the graph, as long as the re- mitters that create an “odd hole” (arrows) Dana Mackenzie is a writer in Santa Cruz, 1 maining nodes keep their links intact. A cell- makes the graph imperfect. California. CREDIT: GRAPH BY COLIN MCDIARMID/UNIVERSITY OF OXFORD FROM ADAPTED 38 5 JULY 2002 VOL 297 SCIENCE www.sciencemag.org.
Recommended publications
  • Lecture 10: April 20, 2005 Perfect Graphs
    Re-revised notes 4-22-2005 10pm CMSC 27400-1/37200-1 Combinatorics and Probability Spring 2005 Lecture 10: April 20, 2005 Instructor: L´aszl´oBabai Scribe: Raghav Kulkarni TA SCHEDULE: TA sessions are held in Ryerson-255, Monday, Tuesday and Thursday 5:30{6:30pm. INSTRUCTOR'S EMAIL: [email protected] TA's EMAIL: [email protected], [email protected] IMPORTANT: Take-home test Friday, April 29, due Monday, May 2, before class. Perfect Graphs k 1=k Shannon capacity of a graph G is: Θ(G) := limk (α(G )) : !1 Exercise 10.1 Show that α(G) χ(G): (G is the complement of G:) ≤ Exercise 10.2 Show that χ(G H) χ(G)χ(H): · ≤ Exercise 10.3 Show that Θ(G) χ(G): ≤ So, α(G) Θ(G) χ(G): ≤ ≤ Definition: G is perfect if for all induced sugraphs H of G, α(H) = χ(H); i. e., the chromatic number is equal to the clique number. Theorem 10.4 (Lov´asz) G is perfect iff G is perfect. (This was open under the name \weak perfect graph conjecture.") Corollary 10.5 If G is perfect then Θ(G) = α(G) = χ(G): Exercise 10.6 (a) Kn is perfect. (b) All bipartite graphs are perfect. Exercise 10.7 Prove: If G is bipartite then G is perfect. Do not use Lov´asz'sTheorem (Theorem 10.4). 1 Lecture 10: April 20, 2005 2 The smallest imperfect (not perfect) graph is C5 : α(C5) = 2; χ(C5) = 3: For k 2, C2k+1 imperfect.
    [Show full text]
  • The Strong Perfect Graph Theorem
    Annals of Mathematics, 164 (2006), 51–229 The strong perfect graph theorem ∗ ∗ By Maria Chudnovsky, Neil Robertson, Paul Seymour, * ∗∗∗ and Robin Thomas Abstract A graph G is perfect if for every induced subgraph H, the chromatic number of H equals the size of the largest complete subgraph of H, and G is Berge if no induced subgraph of G is an odd cycle of length at least five or the complement of one. The “strong perfect graph conjecture” (Berge, 1961) asserts that a graph is perfect if and only if it is Berge. A stronger conjecture was made recently by Conforti, Cornu´ejols and Vuˇskovi´c — that every Berge graph either falls into one of a few basic classes, or admits one of a few kinds of separation (designed so that a minimum counterexample to Berge’s conjecture cannot have either of these properties). In this paper we prove both of these conjectures. 1. Introduction We begin with definitions of some of our terms which may be nonstandard. All graphs in this paper are finite and simple. The complement G of a graph G has the same vertex set as G, and distinct vertices u, v are adjacent in G just when they are not adjacent in G.Ahole of G is an induced subgraph of G which is a cycle of length at least 4. An antihole of G is an induced subgraph of G whose complement is a hole in G. A graph G is Berge if every hole and antihole of G has even length. A clique in G is a subset X of V (G) such that every two members of X are adjacent.
    [Show full text]
  • Research Topics in Graph Theory and Its Applications
    Research Topics in Graph Theory and Its Applications Research Topics in Graph Theory and Its Applications By Vadim Zverovich Research Topics in Graph Theory and Its Applications By Vadim Zverovich This book first published 2019 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2019 by Vadim Zverovich All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-5275-3533-9 ISBN (13): 978-1-5275-3533-6 Disclaimer: Any statements in this book might be fictitious and they represent the author's opinion. In memory of my son, Vladik (1987 { 2000) . Contents Preface xi 1 α-Discrepancy and Strong Perfectness 1 1.1 Background and Aims . 1 1.2 α-Discrepancy . 4 1.3 Strongly Perfect Graphs . 7 1.4 Computational Aspects . 10 1.5 Beneficiaries and Dissemination . 13 1.6 References . 14 1.7 Resources and Work Plan . 16 1.8 Reviewers' Reports . 19 1.9 PI's Response to Reviewers' Comments . 30 1.10 Summary of Key Lessons . 32 2 Braess' Paradox in Transport Networks 35 2.1 Background and Aims . 35 2.2 Braess' Paradox and Methodology . 40 2.3 Computational Aspects of Braess' Paradox . 44 2.4 Braess-like Situations in Rail Networks . 48 2.5 References . 49 2.6 Beneficiaries, Resources and Work Plan .
    [Show full text]
  • A Constructive Formalization of the Weak Perfect Graph Theorem
    A Constructive Formalization of the Weak Perfect Graph Theorem Abhishek Kr Singh1 and Raja Natarajan2 1 Tata Institute of Fundamental Research, Mumbai, India [email protected] 2 Tata Institute of Fundamental Research, Mumbai, India [email protected] Abstract The Perfect Graph Theorems are important results in graph theory describing the relationship between clique number ω(G) and chromatic number χ(G) of a graph G. A graph G is called perfect if χ(H) = ω(H) for every induced subgraph H of G. The Strong Perfect Graph Theorem (SPGT) states that a graph is perfect if and only if it does not contain an odd hole (or an odd anti-hole) as its induced subgraph. The Weak Perfect Graph Theorem (WPGT) states that a graph is perfect if and only if its complement is perfect. In this paper, we present a formal framework for working with finite simple graphs. We model finite simple graphs in the Coq Proof Assistant by representing its vertices as a finite set over a countably infinite domain. We argue that this approach provides a formal framework in which it is convenient to work with different types of graph constructions (or expansions) involved in the proof of the Lovász Replication Lemma (LRL), which is also the key result used in the proof of Weak Perfect Graph Theorem. Finally, we use this setting to develop a constructive formalization of the Weak Perfect Graph Theorem. Digital Object Identifier 10.4230/LIPIcs.CPP.2020. 1 Introduction The chromatic number χ(G) of a graph G is the minimum number of colours needed to colour the vertices of G so that no two adjacent vertices get the same colour.
    [Show full text]
  • Classes of Perfect Graphs
    This paper appeared in: Discrete Mathematics 306 (2006), 2529-2571 Classes of Perfect Graphs Stefan Hougardy Humboldt-Universit¨atzu Berlin Institut f¨urInformatik 10099 Berlin, Germany [email protected] February 28, 2003 revised October 2003, February 2005, and July 2007 Abstract. The Strong Perfect Graph Conjecture, suggested by Claude Berge in 1960, had a major impact on the development of graph theory over the last forty years. It has led to the definitions and study of many new classes of graphs for which the Strong Perfect Graph Conjecture has been verified. Powerful concepts and methods have been developed to prove the Strong Perfect Graph Conjecture for these special cases. In this paper we survey 120 of these classes, list their fundamental algorithmic properties and present all known relations between them. 1 Introduction A graph is called perfect if the chromatic number and the clique number have the same value for each of its induced subgraphs. The notion of perfect graphs was introduced by Berge [6] in 1960. He also conjectured that a graph is perfect if and only if it contains, as an induced subgraph, neither an odd cycle of length at least five nor its complement. This conjecture became known as the Strong Perfect Graph Conjecture and attempts to prove it contributed much to the developement of graph theory in the past forty years. The methods developed and the results proved have their uses also outside the area of perfect graphs. The theory of antiblocking polyhedra developed by Fulkerson [37], and the theory of modular decomposition (which has its origins in a paper of Gallai [39]) are two such examples.
    [Show full text]
  • Strong Perfect Graph Theorem a Graph G Is Perfect If and Only If Neither G Nor Its Complement G¯ Contains an Induced Odd Circuit of Length ≥ 5
    THEOREM OF THE DAY The Strong Perfect Graph Theorem A graph G is perfect if and only if neither G nor its complement G¯ contains an induced odd circuit of length ≥ 5. An induced odd circuit is an odd-length, circular sequence of edges having no ‘short-circuit’ edges across it, while G¯ is the graph obtained by replacing edges in G by non-edges and vice-versa. A perfect graph G is one in which, for every induced subgraph H, the size of a largest clique (that is, maximal complete subgraph) is equal to the chromatic number of H (the least number of vertex colours guaranteeing no identically coloured adjacent vertices). This deep and subtle property is confirmed by today’s theorem to have a surprisingly simple characterisation, whereby the railway above is clearly perfect. The railway scenario illustrates just one way in which perfect graphs are important. We wish to dispatch goods every day from depots v1, v2,..., choosing the best-stocked depots but subject to the constraint that we nominate at most one depot per network clique, so as to avoid head-on collisions. The depot-clique incidence relationship is modelled as a 0-1 matrix and we attempt to replicate our constraint numerically from this as a set of inequalities (far right, bottom). Now we may optimise dispatch as a standard linear programming problem unless... the optimum allocates a fractional amount to each depot, failing to respect the one-depot-per-clique constraint. A 1975 theorem of V. Chv´atal asserts: if a clique incidence matrix is the constraint matrix for a linear programme then an integer optimal solution is guaranteed if and only if the underlying network is a perfect graph.
    [Show full text]
  • Perfect Graphs the American Institute of Mathematics
    Perfect Graphs The American Institute of Mathematics This is a hard{copy version of a web page available through http://www.aimath.org Input on this material is welcomed and can be sent to [email protected] Version: Tue Aug 24 11:37:57 2004 0 The chromatic number of a graph G, denoted by Â(G), is the minimum number of colors needed to color the vertices of G in such a way that no two adjacent vertices receive the same color. Clearly Â(G) is bounded from below by the size of a largest clique in G, denoted by !(G). In 1960, Berge introduced the notion of a perfect graph. A graph G is perfect, if for every induced subgraph H of G, Â(H) = !(H). A hole in a graph is a chordless cycle of length greater than 3, and it is even or odd depending on the number of vertices it contains. An antihole is the complement of a hole. It is easily seen that odd holes and odd antiholes are not perfect. Berge conjectured that these are the only minimal imperfect graphs, i.e., a graph is perfect if and only if it does not contain an odd hole nor an odd antihole. (When we say that a graph G contains a graph H, we mean as an induced subgraph). This was known as the Strong Perfect Graph Conjecture (SPGC), whose proof has been announced recently. 0This document was organized by Maria Chudnovsky as part of the lead-in and followup to the ARCC focused workshop \The Perfect Graph Conjecture," October 29 to Novenber 2, 2002.
    [Show full text]
  • Arxiv:2012.02851V1 [Math.CO]
    Some New Results Concerning Power Graphs and Enhanced Power Graphs of Groups I. Boˇsnjak, R. Madar´asz, S. Zahirovi´c Abstract The directed power graph G~(G) of a group G is the simple digraph with vertex set G such that x → y if y is a power of x. The power graph of G, denoted by G(G), is the underlying simple graph. The enhanced power graph Ge(G) of G is the simple graph with vertex set G in which two elements are adjacent if they generate a cyclic subgroup. In this paper it is proved that, if two groups have isomorphic power graphs, then they have isomorphic enhanced power graphs too. A suffi- cient condition for a group to have perfect enhanced power graph is given. By this condition, any group having order divisible by at most two primes has perfect enhanced power graph. 1 Introduction The directed power graph of a group G is the simple directed graph whose vertex set is G, and in which x → y if y is a power of x. Its underlying simple graph is called the power graph of the group. The directed power graph was introduced by Kelarev and Quinn [15], while the power graph was first studied by Chakrabarty, Ghosh and Sen [11]. The power graph has been subject of many papers, including [1, 6–10, 16–18, 22, 23]. In these papers, combinatorial and algebraic properties of the power graph have received great attention. For more details the survey [2] is recommended. The enhanced power graph of a group is the simple graph whose vertices are elements of the group, and in which two vertices are adjacent if they are powers of some element of the group, i.e.
    [Show full text]
  • Colouring Perfect Graphs with Bounded Clique Number Maria Chudnovsky, Aurélie Lagoutte, Paul Seymour, Sophie Spirkl
    Colouring perfect graphs with bounded clique number Maria Chudnovsky, Aurélie Lagoutte, Paul Seymour, Sophie Spirkl To cite this version: Maria Chudnovsky, Aurélie Lagoutte, Paul Seymour, Sophie Spirkl. Colouring perfect graphs with bounded clique number. Journal of Combinatorial Theory, Series B, Elsevier, 2017, 122, pp.757-775. 10.1016/j.jctb.2016.09.006. hal-01561528 HAL Id: hal-01561528 https://hal.archives-ouvertes.fr/hal-01561528 Submitted on 12 Jul 2017 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Colouring perfect graphs with bounded clique number Maria Chudnovsky1 Princeton University, Princeton, NJ 08544 Aur´elie Lagoutte2 LIP, UMR 5668, ENS Lyon, CNRS, UCBL, INRIA, Universit´ede Lyon, France Paul Seymour3 Princeton University, Princeton, NJ 08544 Sophie Spirkl Princeton University, Princeton, NJ 08544 November 4, 2015; revised September 2, 2016 1Supported by NSF grant DMS-1550991 and US Army Research Office Grant W911NF-16-1-0404. 2Partially supported by ANR Grant STINT:ANR-13-BS02-0007, and performed while visiting Princeton University. 3Supported by ONR grant N00014-14-1-0084 and NSF grant DMS-1265563. Abstract A graph is perfect if the chromatic number of every induced subgraph equals the size of its largest clique, and an algorithm of Gr¨otschel, Lov´asz, and Schrijver [9] from 1988 finds an optimal colouring of a perfect graph in polynomial time.
    [Show full text]
  • On Characterizing Game-Perfect Graphs by Forbidden Induced Subgraphs
    Volume 7, Number 1, Pages 21{34 ISSN 1715-0868 ON CHARACTERIZING GAME-PERFECT GRAPHS BY FORBIDDEN INDUCED SUBGRAPHS STEPHAN DOMINIQUE ANDRES Abstract. A graph G is called g-perfect if, for any induced subgraph H of G, the game chromatic number of H equals the clique number of H. A graph G is called g-col-perfect if, for any induced subgraph H of G, the game coloring number of H equals the clique number of H. In this paper we characterize the classes of g-perfect resp. g-col-perfect graphs by a set of forbidden induced subgraphs. Moreover, we study similar notions for variants of the game chromatic number, namely B-perfect and [A; B]-perfect graphs, and for several variants of the game coloring number, and characterize the classes of these graphs. 1. Introduction A well-known maker-breaker game is one of Bodlaender's graph coloring games [9]. We are given an initially uncolored graph G and a color set C. Two players, Alice and Bob, move alternately with Alice beginning. A move consists in coloring an uncolored vertex with a color from C in such a way that adjacent vertices receive distinct colors. The game ends if no move is possible any more. The maker Alice wins if the vertices of the graph are completely colored, otherwise, i.e. if there is an uncolored vertex surrounded by colored vertices of each color, the breaker Bob wins. For a graph G, the game chromatic number χg(G) of G is the smallest cardinality of a color set C such that Alice has a winning strategy in the game described above.
    [Show full text]
  • Characterizations of Perfect Graphs
    Takustraße 7 Konrad-Zuse-Zentrum D-14195 Berlin-Dahlem f¨ur Informationstechnik Berlin Germany MARTIN GROTSCHEL¨ My Favorite Theorem: Characterizations of Perfect Graphs Preprint SC 99-17 (June 1999) MY FAVORITE THEOREM: CHARACTERIZATIONS OF PERFECT GRAPHS MARTIN GROTSCHEL¨ The favorite topics and results of a researcher change over time, of course. One area that I have always kept an eye on is that of perfect graphs. These graphs, in- troduced in the late 50s and early 60s by Claude Berge, link various mathematical disciplines in a truly unexpected way: graph theory, combinatorial optimization, semidefinite programming, polyhedral and convexity theory, and even informati- on theory. This is not a survey of perfect graphs. It’s just an appetizer. To learn about the origins of perfect graphs, I recommend to read the historical papers [1] and [2]. The book [3] is a collection of important articles on perfect graphs. Algorithmic aspects of perfect graphs are treated in [13]. A comprehensive survey of graph classes, including perfect graphs, can be found in [5]. Hundreds of classes of perfect graphs are known, 96 important classes and the inclusion relations among them are described in [16]. So, what is a perfect graph? Complete graphs are perfect, bipartite, interval, com- parability, triangulated, parity, and unimodular graphs are perfect as well. The following beautiful perfect graph is the line graph of the complete bipartite graph K3,3. Due to the evolution of the theory, definitions of perfection (and versions thereof) have changed over time. To keep this article short, I do not follow the historical 1 development of the notation.
    [Show full text]
  • II. Stable Sets and Colourings
    II. Stable sets and colourings 1. Stable sets and colourings Let G = (V,E) be a graph. A stable set is a subset S of V containing no edge of G. A clique is a subset C of V such that any two vertices in C are adjacent. So (1) S is a stable set of G S is a clique of G, ⇐⇒ where G denotes the complementary graph of G.1 A vertex-colouring or colouring of G is a partition Π of V into stable sets S1,...,Sk. The sets S1,...,Sk are called the colours of the colouring. A clique cover of G is a partition Π of V into cliques. Define: (2) α(G) := max S S is a stable set , {| || } ω(G) := max C C is a clique , {| || } χ(G) := min Π Π is a colouring , {| || } χ(G) := min Π Π is a clique cover . {| || } These numbers are called the stable set number, the clique number, the vertex-colouring number or colouring number, and the clique cover number of G, respectively. We say that a graph G is k-(vertex-)colourable if χ(G) k. ≤ Note that (3) α(G)= ω(G) and χ(G)= χ(G). We have seen that in any graph G = (V,E), a maximum-size matching can be found in polynomial time. This means that α(L(G)) can be found in polynomial time, where L(G) is the line graph of G.2 On the other hand, it is NP-complete to find a maximum-size stable set in a graph.
    [Show full text]