/
Use of MAXCUT for Ramsey Arrowing of Triangles Alexand Use of MAXCUT for Ramsey Arrowing of Triangles Alexand

Use of MAXCUT for Ramsey Arrowing of Triangles Alexand - PDF document

pamella-moone
pamella-moone . @pamella-moone
Follow
405 views
Uploaded On 2015-04-30

Use of MAXCUT for Ramsey Arrowing of Triangles Alexand - PPT Presentation

Lange Stanis law P Radziszowski Department of Computer Science Rochester Institute of Technology Rochester NY 14623 arl9577spr csritedu and Xiaodong Xu Guangxi Academy of Sciences Nanning Guangxi 530007 China xxdmathssinacom Abstract In 1967 Erd733o ID: 57885

Lange Stanis law

Share:

Link:

Embed:

Download Presentation from below link

Download Pdf The PPT/PDF document "Use of MAXCUT for Ramsey Arrowing of Tri..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

1IntroductionGivenasimplegraphG,wewriteG!(a1;:::;ak)eandsaythatGarrows(a1;:::;ak)eifforeveryedgek-coloringofG,amonochromaticKaiisforcedforsomecolori2f1;:::;kg.Likewise,forgraphsFandH,G!(F;H)eifforeveryedge2-coloringofG,amonochromaticFisforcedinthe rstcolororamonochromaticHisforcedinthesecond.De neFe(a1;:::;ak;p)tobethesetofallgraphsthatarrow(a1;:::;ak)eanddonotcontainKp;theyareoftencalledFolkmangraphs.TheedgeFolkmannumberFe(a1;:::;ak;p)isthesmallestorderofagraphthatisamemberofFe(a1;:::;ak;p).In1970,Folkman[6]showedthatfork�maxfs;tg,Fe(s;t;k)exists.TherelatedproblemofvertexFolkmannumbers,whereverticesarecoloredinsteadofedges,ismorestudied[16,18]thanedgeFolkmannumbers,butwewillnotbediscussingthem.Therefore,wewillskiptheuseofthesuperscriptewhendiscussingarrowing,asitisusuallyusedtodistinguishbetweenedgeandvertexcolorings.In1967,Erd}osandHajnal[5]askedthequestion:DoesthereexistaK4-freegraphthatisnottheunionoftwotriangle-freegraphs?ThisquestionisequivalenttoaskingfortheexistenceofaK4-freegraphsuchthatinanyedge2-coloring,amonochromatictriangleisforced.AfterFolkmanprovedtheexistenceofsuchagraph,thequestionthenbecameto ndhowsmallthisgraphcouldbe,orusingtheabovenotation,whatisthevalueofFe(3;3;4).Priortothispaper,thebestknownboundsforthiscasewere19Fe(3;3;4)941[21,4].FolkmannumbersarerelatedtoRamseynumbersR(s;t),whicharede nedastheleastpositivensuchthatany2-coloringoftheedgesofKnyieldsamonochromaticKsinthe rstcolororamonochromaticKtinthesecondcolor.Usingthearrowingoperator,itisclearthatR(s;t)isthesmallestnsuchthatKn!(s;t).TheknownvaluesandboundsforvarioustypesofRamseynumbersarecollectedandregularlyupdatedbythesecondauthor[20].Wewillbeusingstandardgraphtheorynotation:V(G)andE(G)forthevertexandedgesetsofgraphG,respectively.Acutisapartitionoftheverticesofagraphintotwosets,SV(G)and S=V(G)nS.Thesizeofacutisthenumberofedgesthatjointhetwosets,thatis,jffu;vg2E(G)ju2Sandv2 Sgj.MAX-CUTisawell-knownNP-hardcombinatorialoptimizationproblemwhichasksforthemaximumsizeofacutofagraph. smallestorderofanyFolkmangraph(with xedparameters)seemstobedicult,andsomerelatedRamseygraphcoloringproblemsareNP-hardorlieevenhigherinthepolynomialhierarchy.Forexample,Burr[2]showedthatarrowing(3;3)iscoNP-complete,andSchaefer[22]showedthatforgeneralgraphsF,G,andH,F!(G;H)isP2-complete.3ArrowingviaMAX-CUTBuildingo Spencer'sandothermethods,DudekandRodl[4]in2008showedhowtoconstructagraphHGfromagraphG,suchthatthemax-imumsizeofacutofHGdetermineswhetherornotG!(3;3).TheyconstructthegraphHGasfollows.TheverticesofHGaretheedgesofG,sojV(HG)j=jE(G)j.Fore1;e22V(HG),ifedgesfe1;e2;e3gformatriangleinG,thenfe1;e2gisanedgeinHG.Lett4(G)denotethenumberoftrianglesingraphG.Clearly,jE(HG)j=3t4(G).LetMC(H)denotetheMAX-CUTvalueofgraphH.Theorem1(DudekandRodl[4]).G!(3;3)ifandonlyifMC(HG)2t4(G).ThereisaclearintuitionbehindTheorem1thatwewillnowdescribe.Anyedge2-coloringofGcorrespondstoabipartitionoftheverticesinHG.IfatrianglecoloredinGisnotmonochromatic,thenitsthreeedges,whichareverticesofHG,willbeseparatedinthebipartition.Ifwetreatthisbipartitionasacut,thenthesizeofthecutwillcounteachtriangletwiceforthetwoedgesthatcrossit.Sincethereisonlyonetriangleinagraphthatcontainstwogivenedges,thise ectivelycountsthenumberofnon-monochromatictriangles.Therefore,ifitispossibleto ndacutthathassizeequalto2t4(G),thensuchacutde nesanedgecoloringofGthathasnomonochromatictriangles.However,ifMC(HG)2t4(G),thenineachcoloring,allthreeedgesofsometriangleareinonepartandthus,G!(3;3).Abene tofconvertingtheproblemofarrowing(3;3)toMAX-CUTisthatthelatteriswell-knownandhasbeenstudiedextensivelyincomputerscienceandmathematics(seeforexample[3]).ThedecisionproblemMAX-CUT(H;k)askswhetherornotMC(H)k.ItisknownthatMAX-CUTisNP-hardandthisdecisionproblemwasoneofKarp's21NP-completeproblems[13].Inourcase,G!(3;3)ifandonlyifMAX-CUT(HG;2t4(G))doesn'thold.SinceMAX-CUTisNP-hard,anattemptisoftenmadetoapproximateit,suchasintheapproachespresentedinthenexttwosections. 3.2Goemans-WilliamsonMethodTheGoemans-WilliamsonMAX-CUTapproximationalgorithm[9]isawell-known,polynomial-timealgorithmthatrelaxestheproblemtoasemi-de niteprogram(SDP).Itinvolvesthe rstuseofSDPincombinatorialapproximationandhassinceinspiredavarietyofothersuccessfulalgo-rithms(seeforexample[12,8]).Thisrandomizedalgorithmreturnsacutwithexpectedsizeatleast0.87856oftheoptimalvalue.However,inourcase,allthatisneededisafeasiblesolutiontotheSDP,asitgivesanupperboundonMC(H).AbriefdescriptionoftheGoemans-Williamsonrelaxationfollows.The rststepinrelaxingMAX-CUTistorepresenttheproblemasaquadraticintegerprogram.GivenagraphHwithV(H)=f1;:::;ngandnonnegativeweightswi;jforeachpairofverticesfi;jg,wecanwriteMC(H)asthefollowingobjectivefunction:Maximize1 2Xiwi;j(1�yiyj)(3)subjectto:yi2f�1;1gforalli2V(H):De neonepartofthecutasS=fijyi=1g.Sinceinourcaseallgraphsareweightless,wewillusewi;j=(1iffi;jg2E(H);0otherwise:Next,theintegerprogram(3)isrelaxedbyextendingtheproblemtohigherdimensions.Eachyi2f�1;1gisnowreplacedwithavectorontheunitspherevi2Rn,asfollows:Maximize1 2Xiwi;j(1�vivj)(4)subjectto:kvik=1foralli2V(H):Ifwede neamatrixYwiththeentriesyi;j=vivj,thatis,theGrammatrixofv1;:::;vn,thenyi;i=1andYispositivesemi-de nite.Therefore,(4)isasemide niteprogram.3.3SomeCasesofArrowingUsingtheGoemans-Williamsonapproach,wetestedawidevarietyofgraphsforarrowingby ndingupperboundsonMAX-CUT.Thesegraphsincluded G 2t4(G) min SDP L(127;5) 19558 20181 20181 L(457;6) 347320 358204 358204 L(761;3) 694032 731858 731858 L(785;53) 857220 857220 857220 G786 857762 857843 857753 Table2:PotentialFe(3;3;4)graphsGandupperboundsonMC(HG),where\min"isthebound(1)and\SDP"isthesolutionof(4)fromSDPLR-MCandSBmethod.G786isthegraphofTheorem3.oneoftheMAX-CUTmethodswouldbeabletoprovearrowing.Indeed,wewereabletodosowithL(785;53).NoticethatalloftheupperboundsforMC(HL(785;53))are857220,thesameas2t4(L(785;53)).OurgoalwasthentoslightlymodifyL(785;53)sothatthisvaluebecomessmaller.LetG786denotethegraphL(785;53)withoneadditionalvertexconnectedtothefollowing60vertices:{0,1,3,4,6,7,9,10,12,13,15,16,18,19,21,22,24,25,27,28,30,31,33,34,36,37,39,40,42,43,45,46,48,49,51,52,54,55,57,58,60,61,63,66,69,201,204,207,210,213,216,219,222,225,416,419,422,630,642,645}G786isstillK4-free,has61290edges,andhas428881triangles.TheupperboundcomputedfromtheSDPsolversforMC(HG786)is857753.Wedidnot ndanicedescriptionforthevectorsofthissolution.SoftwareimplementingSpeeDPbyGrippoetal.[10],analgorithmdesignedtosolvelargeMAX-CUTSDPrelaxations,wasusedbyRinaldi(oneoftheauthorsof[10])toanalyzethisgraph.Hewasabletoobtainthebounds857742MC(HG786)857750,whichagreeswith,andimprovesoverourupperboundcomputation.Since2t4(G786)=857762,wehavebothfromourtestsandhisSpeeDPtestthatG786!(3;3),andthefollowingmainresult.Theorem3.Fe(3;3;4)786:Wenotethat ndingalowerboundonMAX-CUT,suchasthe857742MC(HG786)boundfromSpeeDP,followsfrom ndinganactualcutofacertainsize.Thismethodmaybeuseful,as ndingacutofsize2t4(G)showsthatG6!(3;3). [7]PeterFranklandVojtechRodl.Largetriangle-freesubgraphsingraphswithoutK4.GraphsandCombinatorics,2:135{144,1986.[8]AlanFriezeandMarkJerrum.ImprovedApproximationAlgorithmsforMAXk-CUTandMAXBISECTION.Algorithmica,18(1):67{81,1997.[9]MichaelGoemansandDavidWilliamson.ImprovedApproxima-tionAlgorithmsforMaximumCutandSatis abilityProblemsUs-ingSemide niteProgramming.JournaloftheACM,42(6):1115{1145,1995.[10]LuigiGrippo,LauraPalagi,MauroPiacentini,VeronicaPiccialli,andGiovanniRinaldi.SpeeDP:AnalgorithmtocomputeSDPboundsforverylargeMax-Cutinstances.MathematicalProgramming,2012.doi:10.1007/s10107-012-0593-0.[11]ChristophHelmbergandFranzRendl.ASpectralBun-dleMethodforSemide niteProgramming.SIAMJournalofOptimization,10:673{696,2000.Softwareavailableathttp://www-user.tu-chemnitz.de/~helmberg.[12]HowardKarlo andUriZwick.A7=8ApproximationAlgorithmforMAX3SAT?In38thAnnualIEEESymposiumonFoundationsofComputerScience,pages406{415,1997.[13]RichardM.Karp.ReducibilityAmongCombinatorialProblems.InR.E.MillerandJ.W.Thatcher,editors,ComplexityofComputerComputations,pages85{103.Plenum,NewYork,1972.[14]ShenLin.OnRamseynumbersandKr-coloringofgraphs.JournalofCombinatorialTheory,SeriesB,12:82{92,1972.[15]LinyuanLu.ExplicitConstructionofSmallFolkmanGraphs.SIAMJournalonDiscreteMathematics,21(4):1053{1060,January2008.[16]Tomasz Luczak,AndrzejRucinski,andSebastianUrbanski.Onmini-malFolkmangraphs.DiscreteMathematics,236:245{262,2001.[17]MATLAB.Version7.12.0(R2011a).TheMathWorksInc.,Natick,Massachusetts,2011.http://www.mathworks.com/products/matlab.[18]NedyalkoNenov.OnthetrianglevertexFolkmannumbers.DiscreteMathematics,271:327{334,September2003.