ebook img

Program - Alex Grasas PDF

351 Pages·2010·2.17 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Program - Alex Grasas

T S ECHNICAL ESSIONS Monday, 9:00-10:20 We consider the NP-hard Steiner tree problem under a two-stage stochastic modelwithrecourseandfinitelymanyscenarios.Wediscussundirected,semi- directed, and directed cut-set based integer programming models, and sug- (cid:4) gestabranch-and-cutapproachbasedonBendersdecompositionandaderived MA-01 Integer-L-shapedalgorithm. Wecompareourdifferentmodelsboththeoreti- Monday,9:00-10:20 cally,namelyfromapolyhedralpointofview,andcomputationally. AulaMagna 3- ExactandHeuristicAlgorithmsfortheBoundedCycle Keynote Talk 1 CoverProblem Stream: KeynoteSpeakers IreneLoiseau,DepartamentodeComputación-,Facultadde CienciasExactasyNaturales,UniversidaddeBuenosAires, Invitedsession PabellónI-CiudadUniversitaria,1428,BuenosAires, Chair:Gerhard-WilhelmWeber,InstituteofAppliedMathematics, Argentina,[email protected] MiddleEastTechnicalUniversity,ODTÜ,06531,Ankara,Turkey, [email protected] WewillpresentanewexactalgorithmandheuristicsfortheBoundedCycle CoverProblem. BCCPrequirestodetermineaminimumcostcyclecoverof agraph,withcyclesboundedinlengthandnumberofedges. Thisproblem 1- Safetractableapproximationsofchanceconstraints arisesinseveralsituationsrelatedtotelecommunicationsnetworksdesign,as, forexample,whenwewanttodesignfiber-optictelecommunicationsnetworks ArkadiNemirovski,SchoolofIndustrialandSystems thatemploymultipleself-healingringstoprovideroutingforcommunication Engineering,GeorgiaInstituteofTechnology,765FerstDrive, traffic,evenintheeventofafibercutorotherkindsoffailures. NW,GA30332-0205,Atlanta,Germany, [email protected] 4- Reverse Multistar Inequalities and Vehicle Routing Whenoptimizingunderstochasticuncertainty,theentityofprimaryimportance Problemswithlowerboundcapacities isachanceconstraintProbqsi->P f(x;qsi)inQ >=1-epsilon,forallPinPP wherexisthedecisionvector,qsiisarandomperturbationwithdistributionP LuisGouveia,DEIO,UniversityofLisbon,CampoGrande, knowntobelongtoagivenfamilyPP,Qisagiventargetset,andepsilon«1 BlocoC6,1749-016,Lisbon,Portugal,[email protected],Juan isagiventolerance.Asideofahandfulofspecialcases,chanceconstrainsare JoséSalazarGonzález computationallyintractable:first,itisdifficulttocheckefficientlywhetherthe In this talk we discuss, present and test models for the Capacitated Vehicle constraintissatisfiedatagivenx,andsecond,thefeasiblesetofachancecon- RoutingProblemwitharclowerbounds. Weintroducetheso-calledreversed strainttypicallyisnonconvex,whichmakesitproblematictooptimizeunder multistarinequalitiesandtwootherrelatedfamiliesofinequalities,andshow theconstraint. Giventhesedifficulties,anaturalwaytoprocessachancecon- thattheyarerelevantformodellingthisroutingproblem. Wepresentresults straintistoreplaceitwithitssafetractableapproximationatractableconvex fromabranch-and-cutalgorithmwhichusesthenewinequalitiesforsolving constraintwiththefeasiblesetcontainedintheoneofthechanceconstraint. instanceswithupto50customers. Inthetalk,weoverviewsomerecentresultsinthisdirection,withemphasison chanceversionsofwell-structuredconvexconstraints(primarily,affinelyper- turbedscalarlinearandlinearmatrixinequalities)andestablishlinksbetween thistopicandRobustOptimization. (cid:4) MA-03 Monday,9:00-10:20 (cid:4) MA-02 3.2.15 Monday,9:00-10:20 TSP 3.2.14 Stream: Metaheuristics Combinatorial Optimization I Invitedsession Chair:JoséA.Moreno-Pérez,Estadística,I.O.yComputación, Stream: CombinatorialOptimization UniversityofLaLaguna,LaLaguna,Spain,[email protected] Invitedsession Chair:NorainiMohdRazali,Mechanical&Manufacturing Chair:IvanaLjubic,DepartmentofStatisticsandDecisionSupport Engineering,DublinCityUniversity,Dublin9,Dublin,Ireland, Systems,UniversityofVienna,Bruennerstr.72,1210,Vienna, [email protected] Austria,[email protected] 1- An algorithm for solving the traveling salesman prob- 1- Spanning Trees with Node Degree Dependent Costs lem (tsp) based on multimodal transport, using a se- andKnapsackReformulations cuenceoflinearproblems(lp) PedroMoura,DEIO,UniversityofLisbon,CampoGrande, LuisMoreno,Systems,UniversidadNacionaldeColombia, BlocoC6,1749-016,Lisbon,Portugal,[email protected],Luis FacultaddeMinascra.8065-223,BloqueM8oficina207,1, Gouveia Medellin,Antioquia,Colombia,[email protected],Javier TheDegreeconstrainedMinimumSpanningTreeProblem(DMSTP)consists Diaz,GloriaPena infindingaminimalcostspanningtreesatisfyingtheconditionthateverynode Ahybridalgorithmisproposedbasedonheuristicsandlinearprogramming.A hasdegreenogreaterthanafixedvalue. Weconsiderageneralizationofthe relaxedlinearproblem(LP)issolved.Ifthesolutionhasseveralnonconnected DMSTPwithamoregeneralobjectivefunctionincludingmodularcostsasso- circuitsanewheuristicLPisusedtoconnectthecircuits,assumingthecities ciatedtothedegreeofeachnode, whichhaveatechnologicalmotivationin ineachcircuitintherelaxedsolutionareclosebetweenthem.Thecircuitsare thecontextoftelecommunicationsnetworks. WepresentLPmodelstogether connectedaswoulddosomebodyaftervisitingbycarseveralcitiesinaregion withsomevalidinequalitiesandcomparetheirrespectivelinearprogramming thattakesaplanetovisitoptimallyanothersetofregionswithadditionalcities. relaxationsusingcableandwirelessnetworkbasedinstances. Ifwhenconnectingthecircuitofcircuitsthereareagainseveralnonconnected circuitstheheuristicLPissolvedinaniterativewaythatdecreasesthelinear 2- SolvingtheStochasticSteinerTreeProblembyBranch- problemsizeuntilallthecitiesbecomeshortlyandeasilyconnected. and-Cut 2- AModifiedElectromagnetism-likeAlgorithmforTravel- BerndZey,TUDortmund,FrankfurterWeg3,59439, lingSalesmanProblemswithPrecedenceConstraints Holzwickede,Germany,[email protected],Immanuel Bomze,MarkusChimani,MichaelJuenger,IvanaLjubic,Petra AlkinYurtkuran,IndustrialEngineeringDepartment,Uludag Mutzel University,UludagUniversity,IndustrialEngineering 1 MA-04 EURO24-Lisbon2010 WedevelopaTabuSearch(TS)algorithmwithaprobabilisticlocalsearchafter Department,16059,Bursa,Turkey,[email protected],Erdal eachiterationtosolvetheorderacceptanceandschedulingproblemonasingle Emel machinewithsequencedependentsetuptimes. Wecomparetheperformance TravelingSalesmanProblemwithPrecedenceConstraints(TSPPC)isanim- oftheTSalgorithmtoagreedyconstructiveheuristicfromtheliterature,using portantvariantofTravelingSalesmanProblem.TSPPCbelongstotheclassof upperboundsbasedonamixedintegerprogrammingformulation. Computa- NP-Hardproblemswherethereexitsaprecedencerelationshipbetweencus- tionalstudiesshowthattheTSalgorithmgivessignificantlybettersolutions tomers. Inthisstudy,aModifiedElectromagnetism-likeAlgorithm(EMA)is thanthoseoftheconstructiveheuristicintermsofobjectivefunctionvaluein appliedtosolveTSPPCproblems.Thekeyconceptoftheproposedalgorithm allinstancestestedwithasmallincreaseinruntime. istheprojectionoftheparticlespaceontoanewcoordinatespacewhereeach precedenceconstraintisensured. ThecomputationalresultsshowthatModi- 3- Approximate methods for solving the operating room fiedEMAgivespromisingresultswithinacceptablecomputationaltimes. planningproblem 3- Performance comparison between different GA selec- tionstrategiesinsolvingTSPinstance. JoseM.Molina-Pariente,UniversityofSeville,Spain, [email protected],JoseMFraminan,PazPerezGonzalez NorainiMohdRazali,Mechanical&Manufacturing Engineering,DublinCityUniversity,Dublin9,Dublin,Ireland, Inthiscommunication,weaddresstheoperatingroomplanningproblemfora [email protected],JohnGeraghty surgeryunit. Thisprobleminvolvesdeterminingasurgeryschedulethatspec- ifiesthenumberofsurgicalcasestobescheduledinagivenplanninghorizon Thisstudypresentsthecomparisonofgeneticalgorithmperformanceonsolv- togetherwiththedateoftheinterventionandthespecificoperatingroomin ingTSPusingtwodifferentstochasticselectionmethodswhicharetournament whicheachsurgicalcasewillbeperformed.Tosolvethisproblem,wepresent androulettewheel. TheGAismainlycomposedofthreegeneticoperations newconstructiveheuristicsandalocalsearchmethod. Weevaluatetheper- which are selection, crossover and mutation. With the same crossover and formanceofthesemethodsbygeneratingalargesetofinstancesbasedonan mutation operation, the studyis focussedon comparing theeffect of differ- analysisoftheliterature. entselectionstrategytotheperformanceofconvergencethatgivesoptimum solution.NumericalexperimentsshowthatGAwithtournamentselectioncon- vergesmuchfasterthanroulettewheelselection. 4- A GRASP approach to the Multi-Task Employee TimetablingProblem 4- A Data Mining Based Heuristic Approach for Solution ofTravellingSalesmanProblem PilarTormos,STATISTICSANDOPERATIONSRESEARCH, HandeGulkac,ComputerEngineering,OkanUniversity,Turkey, UNIVERSIDADPOLITECNICADEVALENCIA,CAMINO [email protected],SemiyeGönülol,AhmetCihan, DEVERAS/N,VALENCIA,46022,VALENCIA,SPAIN, HalenurS¸ahin,AlpaslanFiglali Spain,[email protected],AntonioLova Travelling Salesman Problem (TSP) is a well known NP-hard optimization EmployeeTimetablingProblem(ETP)istheoperationofassigningemployees problem. Manymethodsareappliedincludingheuristics,mathematicalpro- totasksinasetofshiftsduringaperiodoftimewhilesatisfyingtheexisting grammingandmetaheuristicsforobtaininggoodsolutions.Inthisstudyadata constraintsandpreferences.Anextensionofthisproblem,theMulti-TaskEm- miningbasedheuristicapproachisappliedforthesolutionofTSP.Procedure ployeeTimetablingProblem(MTETP)impliestheassignmentofthesequence usesarepetitiveandforcedrandomtourgenerationapproach. Thesetofbest of tasks to be performed by each employee every working day of the plan- randomtoursareanalyzedviadataminingtoolstoobtaintherelations.Using ninghorizonanditisespeciallyrelevantforcommercialcompanies.Agreedy therulesderivedfromtherelations,theTSPtourisobtained. Theresultsare randomizedadaptivesearchprocedure(GRASP)isdevelopedtosolveitand promisingwhileconsideringit’ssimplicity. embeddedinacomputer-aidedsystem(OPTIHPER).Acustomizedversionof itisinusewithverysatisfactoryresultsbyaleadingSpanishdistributioncom- pany. (cid:4) MA-04 Monday,9:00-10:20 3.2.13 (cid:4) MA-05 Scheduling with metaheuristics Monday,9:00-10:20 3.2.16 Stream: Metaheuristics Invitedsession Theory Chair:LionelAmodeo,CharlesDelaunayInstitute,Universityof Stream: Metaheuristics TechnologyofTroyes,12RueMarieCurieBP2060,10000,Troyes, Invitedsession France,[email protected] Chair:GustavoMelo,ComputerScience,UniversidadeEstadualdo Chair:FaroukYalaoui,InstitutCharlesDelaunay,ICDLOSI, Ceará,Ruacésarfonseca410ap301bairropapicu,60176-110, UniversityofTechnologyofTroyes,12,ruemariecurieBP2060, Fortaleza,Ceará,Brazil,[email protected] 10000,Troyes,France,[email protected] Chair:ZahiraBenkhellat,informatique,sciencesexactes,Bejaia 1- SingleMachineSchedulingwithRejection: Minimizing universityQlgerie,teacher,Algeria,[email protected] totalWeightedCompletionTimeandRejectionCost AtefehMoghaddam,CharlesDelaunayInstitute,Universityof 1- AnalysisofsoftwarefortheNGStechnology: TheSur- TechnologyofTroyes,12rueMarieCurie„Troyes,France, vivalGuide 10000,Troyes,France,[email protected],Farouk Yalaoui,LionelAmodeo BrunoVieira,ComputationalBiology&PopulationGenomics Group,CentrodeBiologiaAmbiental,Departamentode Itisalwaysassumedthatwehavetoprocessalljobs. However,wecanbreak theassumptionbyrejectingcertainjobs. Inthispaper, weconsiderthatthe BiologiaAnimal,FaculdadedeCiênciasdaUniversidadede jobscanbeeitherscheduledonasinglemachineorberejectedatthecostof Lisboa,CampoGrande,1749-016,Lisbon,Portugal, apenalty. Twoobjectivefunctionsareconsidered: minimizingtotalweighted [email protected],FranciscoPinaMartins,SofiaSeabra, completiontimesandminimizingtotalpenalties.Weapplytwo-phasemethod OctavioPaulo tofindallPareto-optimalsolutions. Wealsoproposebi-objectivesimulated annealingalgorithmtofindestimatedPareto-optimalsolutions.Bycomparing NextGenerationSequencing(NGS)technologiesallowthegenerationoflarge thesolutions,weshowthattheresultsarereasonablygood. amountsofdatainashorttimespanandforarelativelylowcost. Amulti- tudeofsoftwarewasrecentlydevelopedtoaddressthedifficultiesgenerated 2- A Tabu Search Algorithm for Order Acceptance and byNGS,suchasassemblingthemillionsofreads,contiggenerationandthe Scheduling followupannotation.Thecoverrateisalsorelevanttotheresultsaccuracyand thedetectionofgeneticvariation,eitherintheformofSNPsorCNVs.Inthis BahriyeCesaret,IndustrialEngineering,KocUniversity,Koc communicationweaddressseveralcriticalbioinformaticsstepsandcompare University,RumelifeneriyoluSariyer,34450,Istanbul,Turkey, currentsoftwareandalgorithmstotackletheseproblems. [email protected],CeydaOguz,SibelSalman 2 EURO24-Lisbon2010 MA-07 (cid:4) (cid:4) MA-06 MA-07 Monday,9:00-10:20 Monday,9:00-10:20 8.2.30 8.2.47 DEA Methodology I Recent Developments in Mathematical Programming Stream: DEAandPerformanceMeasurement Invitedsession Stream: MathematicalProgramming[c] Chair:DimitrisDespotis,DepartmentofInformatics,Universityof Contributedsession Piraeus,80,Karaoli&DimitriouStreet,18534,Piraeus,Greece, Chair:Gerhard-WilhelmWeber,InstituteofAppliedMathematics, [email protected] MiddleEastTechnicalUniversity,ODTÜ,06531,Ankara,Turkey, [email protected] 1- DEAapproachforevaluatingperformanceconsidering Chair:JerzyFilar,MathematicsandStatistics,UniversityofSouth institutiongoal Australia,MawsonLakesBlvd,5095,MawsonLakes,SA,Australia, j.fi[email protected] Sheu-huaChen,DistributionManegement,NationalChin-Yi UniversityofTechnology,Taipin,Taichung,Taiwan,ROC,411, 1- A Relax-and-Fix Lagrangean Relaxation Based Algo- Taichung,Taiwan,Taiwan,[email protected],HongTauLee rithmforaClassofMultiple-ChoiceIntegerProblems Inthedataenvelopmentanalysis(DEA)approachsomedecision-makingunits AbdelkaderSbihi,InformationSystemsandDecisionMaking mayreachperformanceefficiencybytheiroutstandingperformanceonsome Science,Audencia-NantesSchoolofManagement,8routedela relativeunimportantoutputitems. Inthisresearch,wetrytoaddthestrictly Jonelière,BP31222,44312,NantesCedex3,France, predefinedrelationshipsofoutputitemsintheexistedmodelthatexpressedthe [email protected] relativeimportanceofthoseinputoroutputitems.Accordingtothisapproach, onlytheDMUsthatreallymatchthepredefinedrequirementsandhavegood WeproposeaLagrangeanrelaxationbased-algorithmforsomehardcombina- performancecanberegardedasefficientunits. Thatistosaythemeaningof torialproblems. Theideaistorelaxacertaintypeofconstraintsthentofix performancedependsonthegoalstheorganizationpursues.Thisismeaningful variablestotheiroptimalvalue. Weconsider: (i)theMultiple-ChoiceKnap- formanagerialpractices. Acaseofperformanceevaluationoffacultyindif- sackProblem(MCKP)and(ii)theMultiple-ChoiceSubsetSumProblem(MC- ferenttypesofuniversitywithspecificdevelopmentorientationisprovidedto SSP)whichcanbeconsideredasaspecialcaseoftheMCKP.WeusedMCSSP illustratetheproposedidea. asanauxiliaryproblemtotightenthecapacityconstraint. Theobtainedre- sultsshowedahighqualityofthecomputedupperbounds.Thebenchmarkhas demonstratedahighefficiencyoftheapproach. 2- Assessing robustness in additive DEA with interval measurements 2- AnOptimizationMethodforSolvingAssemlyLineBal- ancingProblem MariaGouveia,ISCAC,QuintaAgricola-Bencanta,3040-316, Coimbra,Portugal,[email protected],LuisC.Dias,Carlos SukranSeker,IndustrialEngineeringDepartment,Yildiz HenggelerAntunes TechnicalUniversity,BarborosStreetYildizTechnical UniversityIndustrialEngineeringDepartment343409Besiktas, Thisstudyaddressestheproblemoffindingtherangeofefficiencyforeach Istanbul,Turkey,[email protected],MesutÖzgürler DecisionMakingUnit(DMU)consideringuncertaindata. Uncertaintyinthe DMUcoefficientsineachfactor(inputoroutput)iscapturedthroughinterval Assembly line balancing or simply line balancing is the problem of assign- coefficients(i.e. theseareuncertainbutbounded). Atwo-phaseadditiveData ingoperationstoworkstationsalonganassemblylineinsuchawaythatthe EnvelopmentAnalysis(DEA)modelforperformanceevaluationisused,which assignmentbeoptimalinsomesense. Assigntaskstoworkstationsobserv- isadaptedtoincludetheconceptofsuper-efficiencytoprovidearobustness ingbalancingrestrictionsoastominimizebalancedelaywhilekeepingstation analysisoftheDMUsinfaceofuncertaininformation. workcontentforeverystationcycletime. Therehavebeenalargenumber ofproposalsfortheoreticalandpracticalmethodsforsolvingthelinebalance problem. Thispaperuseoneoftheoptimizationsolutionapproachtosolve 3- A proposition of the minimum distance model in Net- assemblylinebalancingproblem. workDEA 3- Geo-spatial data mining by model-based clustering TohruUeda,FacultyofScienceandTechnology,Seikei methods University,3-3-1Kichijoji-Kitamachi,180-8633, FranciscoFigueiredo,UNIDE,ISCTE-IUL,Lisbon,Portugal, Musashino-Shi,Tokyo,Japan,[email protected],Hirofumi francisco.m.fi[email protected],JoséG.Dias Amatatsu Mostoftheclusteringtechniquesareinadequateforgeo-spatialdataminingas MinimizationofobjectivefunctioninSBMmodelresultsinmaximizationof theytendtoignorethatspatiallycloserareastendtobemoresimilarthanthe slackssum.Thismaximizationcorrespondstofindingapointintheproduction others. Geo-spatialclusteringaimstofindgroupsofsimilarobjectsthatare possibilitysetthatisthefarthestpointfromeachDecisionMakingUnittobe spatiallyclose. Weproposeaclusteringalgorithmforspatialsegmentationof evaluated. Toovercomethisshortage,weproposedtheunifiedDEAmodel. countdataunderaregressionframework,whichcombinestheNeighborhood Traditionallinkingconstraintswherecontinuitybetweeninputandoutputis EM(NEM)andHybridEM(HEM)algorithms. Thegeo-spatialdatamining keptmaybetooseveretoevaluateefficiencies.Consideringlinkingconstraints approachisillustratedwithgeoreferencedpoliticaldata. andtheunifiedDEAmodelwithminimumdistance,weproposeanewnetwork DEAmodelanddiscussefficienciesofprefecturesinJapan. 4- Alternating Proximal Algorithms and Hierarchical Se- lectionofOptimainGames,ControlandPDE’s 4- Animprovingapproachforestimatingreturntoscalein JuanPeypouquet,Mathematics,UniversidadTecnicaFederico DEA SantaMaria,AvEspana1680,2340000,Valparaiso,Valparaiso, Chile,[email protected],HedyAttouch,Marco MaryamAllahyar,mathematics,scienceandresearchbranch Czarnecki islamicazaduniversity,tehran-ashrafiesfahanihighway-to Westudyalternatinganddiagonalproximalalgorithmscombiningresolvent hesarak,0098,tehran,tehran,Iran,IslamicRepublicOf, iterationsandapenalizationscheme. Theresultingsequenceofiteratesand, [email protected],MohsenRostamy-malkhalifeh underlessrestrictiveconditions,theiraveragesconvergeweaklytoapointwith specialproperties.Wealsoanalizeasplittingmethodforstructuredvariational In this article a new method will be suggested for the determination of the problemsandcommentontherobustnessofthesemethods.Theresultsenable rightandleftreturntoscale(RTS).Thenewapproachisdifferentformthatof ustosolveconstrainedorbileveloptimizationproblems. Thismethodisap- GolanyandYu(1997)anddoesn’thaveitsshortcomings.Ourapproachisable pliedtobestresponsedynamicswithcosttochange,optimalcontrolproblems toevaluatetherightandleftRTSinallconditionsforanyunit. anddomaindecompositionforPDE’s. 3 MA-08 EURO24-Lisbon2010 (cid:4) (cid:4) MA-08 MA-09 Monday,9:00-10:20 Monday,9:00-10:20 6.1.36 6.2.53 Project Management Software and Challenges of Mathematical Programming Applications by Modern Applications Stream: ProjectManagementandScheduling Stream: MathematicalProgramming Invitedsession Invitedsession Chair:NorbertTrautmann,DepartmentofBusinessAdministration, Chair:ZuzanaOplatkova,Dept.ofAppliedInformatics,TomasBata UniversityofBern,IFM,APQuantitativeMethoden, UniversityinZlin,NadStranemi4511,76005,Zlin,CzechRepublic, Schützenmattstrasse14,3012,Bern,BE,Switzerland, [email protected] [email protected] Chair:Gerhard-WilhelmWeber,InstituteofAppliedMathematics, Chair:ChristophSchwindt,InstituteofManagementandEconomics, MiddleEastTechnicalUniversity,ODTÜ,06531,Ankara,Turkey, ClausthalUniversityofTechnology,Julius-Albert-Str.2,38678, [email protected] Clausthal-Zellerfeld,Germany,[email protected] 1- An optimization approach for prediction of microbial 1- HeuristicimprovementofMicrosoftProject’sresource- growthstrategies allocationcapabilities PinarOzturk,SystemsBioinformatics,VrijeUniversity, NorbertTrautmann,DepartmentofBusinessAdministration, Amsterdam,DeBoelelaan1085,1081HV,Amsterdam, UniversityofBern,IFM,APQuantitativeMethoden, Netherlands,[email protected],DouweMolenaar Schützenmattstrasse14,3012,Bern,BE,Switzerland, In limited nutrient conditions, microorganisms regulate cellular activities to [email protected],PhilippBaumann maintain efficient growth. Efficiency is regarded as correlated with fast re- MicrosoftProjectiswidelyusedfortemporalschedulingandresourceallo- production. Viaanoptimizationmodelofthewholecellwithanobjectiveto cationofprojects. Weshowthatitsintegratedresource-allocationprocedure maximizegrowthrate,wepredictmacrofeaturesofmicrobesatgivennutri- usesneithertheserialnortheparallelschedule-generationscheme, andthat entconcentrationsbyrepresentingessentialpathwayswithmodules. This,as theprocedureperformsrelativelypoor. Wepresentabi-directionalschedule- knowntous,isthefirsttimegrowthrate,sizeandshapearepredictedhaving improvementheuristic. Computationalresultsforaconstructionprojectand onlyintrinsicphysicalpropertiesofmoleculesasconstraints. fortheJ30,J60,andJ120instancesfromPSPLIBindicatethatthisheuristic shortenstheprojectdurationconsiderably. 2- Robustmodeldevelopmentfornon-linearmodels 2- Exactmethodsforresourcelevellingproblems MagderievanderWesthuizen,SchoolofComputer,Statistical andMathematicalSciences,North-WestUniversity,PrivateBag JürgenZimmermann,OperationsResearch,TUClausthal, X6001,2520,Potchefstroom,SouthAfrica, Julius-AlbertStr.2,38678,Clausthal-Zellerfeld,Germany, [email protected],GielHattingh,Hennie [email protected],ThorstenGather Kruger Wepresentexactsolutionmethodsforresourcelevellingproblemswithmin- imumandmaximumtimelagsamongtheprojectactivities. Inparticular,we Thepredictivecapabilityofregressionmodelsreliesheavilyontheapplicabil- consideratimewindowbasedenumerationmethodandtwotree-basedbranch- ityoftheassumptionsmadebythemodelbuilder. Inaddition,thepresence and-boundproceduresbothwithasophisticatedconstructivelowerbound.Fur- ofoutliersmayalsoleadtomodelsthatarenotreliable. Thisstudyreports thermore,weproposeamixedintegerlinearprogrammingformulationthatcan onrobusttechniquesappliedtominimalassumptionregressionmodelsinan besolvedbystandardsolverssuchasCPLEX.Allapproachesarecomparedin efforttoimprovepredictivecapability. Theapproachisbasedonmathemat- acomprehensivecomputationalstudyusingwellknowntestsetsfromlitera- icalprogrammingtechniquescombinedwithsmoothingandpiecewiselinear ture.Instanceswithupto30activitiescouldbesolvedtooptimality. techniques.Differentcasesfromtheliteratureareconsideredandpresentedas illustrativeexamples. 3- Temporal scheduling of concurrent engineering 3- A Bilevel Competitive Facility Location Model with projects Competitor’sResponse ChristophSchwindt,InstituteofManagementandEconomics, ClausthalUniversityofTechnology,Julius-Albert-Str.2,38678, HandeKucukaydin,IndustrialEngineering,BogaziciUniversity, Clausthal-Zellerfeld,Germany, BogaziciUniversityIndustrialEngineeringDepartment, [email protected],PhilippBenke Bebek-Istanbul-Turkey,34342,Istanbul,Turkey, [email protected],NecatiAras,I.KubanAltinel The concurrent engineering approach is intended to shorten the cycle time of development projects by parallelizing consecutive development phases. We are concerned with a problem in which a new entrant leader firm aims We consider the tradeoff between the time savings enabled by overlapping atfindingthelocationandattractivenessofeachnewfacilitytomaximizeits precedence-relatedprojectactivitiesandtheincreaseintheactivitydurations profitwherethereareexistingfacilitiesbelongingtoacompetitor. Thecom- thatistypicallyincurredbyadditionalintegrationandcoordinationefforts.We petitorreactstotheleaderbyadjustingtheattractivenesslevelsofitsexisting investigatestructuralpropertiesofthetemporalschedulingproblemandex- facilitiestomaximizeitsprofit.Wefirstformulateabilevelmixed-integernon- plainhowearliestandlateststartandcompletiontimesoftheactivitiescanbe linearprogrammingmodel.Then,weconvertitintoanequivalentsinglelevel determinedefficientlybasedonlabel-correctingalgorithms. mixed-integernonlinearprogramandsolveitusingglobaloptimizationmeth- ods. 4- IntegratedSchedulingandStaffingIT-Projects RainerKolisch,TUMSchoolofManagement,Technische (cid:4) UniversitaetMuenchen,Arcisstr.21,80333,Muenchen, MA-10 Germany,[email protected],ChristianHeimerl Monday,9:00-10:20 In this paper we present an optimization model to address the problem of 6.2.56 schedulingtheactivitiesofmultipleIT-projectswithserialstructuresandas- signingtheprojectworktomulti-skilledinternalandexternalhumanresources Graphs and Networks I withstaticandheterogeneousefficiencies. Themixed-binarylinearprogram issolvedusingILOGCPLEXandahybridmetaheuristic. Thelatteremploys Stream: GraphsandNetworks anefficientevaluationfunctionexploitingthenetworkstructureofthestaffing Invitedsession subproblem. Weassesstheimpactsofseveralproblemparametersoncompu- tationtimeandsolutiongaps. Chair:ReinhardtEuler,Informatique,UniversitédeBrest,20av.Le Gorgeu,BP817,29285,Brest,France,[email protected] 4 EURO24-Lisbon2010 MA-11 Statutoryinfectiousdiseasesbreakoutnotonlyaffectspeople’shealthandlives, 1- OnthePowerofDecompositionfortheMaximumInde- butalsostagnatestheeconomicgrowth.Theprevalenceofinfectiousdiseases pendentSetProblem also provides the development opportunities of biotechnology corporations. TheeffectofthestatutoryinfectiousdiseasesoutbreakonTaiwanesebiotech- AndreasBrandstädt,UniversitätRostock,18055,Rostock, nologystockpricemovementsisexamined.Theempiricalresultspointoutthat Germany,[email protected] thereexistsasignificantlypositiveabnormalreturnofTaiwan’sbiotechnology (jointworkwithC.T.Hoang,V.B.Le,V.V.Lozin,andR.Mosca)Inafiniteundi- industrybecauseofthestatutoryinfectiousepidemic. rectedgraphG=(V,E)avertexsetSis’independent’(or’stable’)ifthevertices inSaremutuallynonadjacent.ForgivenG,theMAXIMUMINDEPENDENT SET(MIS)ProblemasksfoanindependentvertexsetofmaximumsizeinG. 2- Ahybridclassificationmethod: Usingasupportvector TheMWSproblemasksforanindependentsetwithmaximumvertexweight; machineforruleextractionondiabetesdiagnosis theMISproblemistheMWSproblemwithunitweights. Itiswellknown thatMWS(MIS,respectively)isintractableandhardtoapproximate. Wedis- Chien-hsinYang,DepartmentofIndustrialEngineeringand cussvariousdecompositiontechniquessuchascliqueseparatordecomposition, Management,OverseasChineseUniversity,100,ChiaoKwang modulardecompositionandsplitdecompositionforsolvingtheMWSprob- Rd.,Taichung40721,Taiwan,Taiwan, lemefficientlyonvariousparticulargraphclasses. Itiswellknownthatthe [email protected],Chun-ChinHsu abovementioneddecompositionsarehelpfultoolsforsolvingtheMWSprob- lem.Oneofourresultsallowstocombinecliqueseparatordecompositionand modulardecomposition.Thisimpliesvariousimprovementsofknownresults, Manyofthefactorsrelatedtodiabetesmellitus(DM)havebeendiscovered amongthemapolynomialtimealgorithmforMWSonapple-freegraphswhich fromasuccessionofstudiesinepidemiology. However,itseemsthattheus- areacommongeneralizationofchordalgraphs,cographsandclaw-freegraphs. abilityandexplainabilityofmethodsareinferiortorulesextraction.Asupport Finallywementionsomeopenproblems. vectormachinebasedhybridclassificationmethodisemployedtoextractrules forDMdiagnosis. Toevaluateperformance,C5andback-propagationneural networkswereusedasbenchmarks.Resultsfromthehybridapproachdemon- 2- Solving efficiently the weighted stable-set problem in stratehighaccuracyandfidelity,andtherulescanhelpforpreventivemedicine claw-freegraphsusingareductionoperation inDMdiagnosis. PaoloNobili,Mathematics,UniversityofLecce,ViaArnesano, 73100,Lecce,Italy,[email protected],AntonioSassano 3- Applicationofartificialneuralnetworkandsupportvec- Maximumweightstablesetscanbecomputedinpolynomialtimeforclaw-free tormachinetoclassifytheriskofdeathofhospitalized graphs(Minty,Nakamuraetal.,Schrijver,Orioloetal.).Inthispaperwedefine patientswithacutecoronarysyndrome thestronglyreduciblecliques,extendingtotheweightedcaseareductionop- erationofLovaszandPlummer.Weusetheoperationforobtainingmaximum RodrigoCollazo,OperationalResearch,CASNAV/UFRJ,Costa weightalternatingpathsthroughmatchingcomputations. Weembedthepro- DoriaSt,17,21910-170,RiodeJaneiro,RiodeJaneiro,Brazil, cedureinaniterativeapproachthatcontructsagivenclaw-freegraphGnode bynode,maintainingtheassociatedmaximumweightstableset.Theresulting [email protected],BasílioPereira,LauraBahiense, algorithmhascomputationalcomplexityO(n4log(n)). AmáliaFariadosReis,AmáliaFariadosReis 3- ReconstructionofPermutationsRespecttosomeGen- Thisstudydevelopedanartificialneuralnetworkmodelandasupportvector machinemodeltoclassifytheriskofdeathofhospitalizedpatientswithacute eratorSetsoftheSymmetricGroup coronarysyndromeathighandlow.Itwasusedthemutualinformationfeature AlparVajkKramer,DEIO,FCUL,Portugal,[email protected] selectorunderuniforminformationdistributioncriteria(MIFS-U)forselection ofthemostimportantinputvariables.Thecomputationalresultsshowabetter We will consider the reconstruction of permutations regarding some special performanceofthesupportvectormachinemodelcomparedwiththeartificial generatorsetsofthesymmetricgroup. Thegeneratorsetsconsideredarepar- neuralnetworkmodelandindicatetheinputvariablesage,creatinineandany ticularsubsetsofinvolutionssuchasthereversals,prefixreversals,bubblere- priorrevascularizationasthemostrelevant. versalsorCoxetergeneratorsandtranspositions. Thecommonpropertyofall thisgeneratorsetsisthattheircorrespondingCayleygraphdoesnotcontain triangles. 4- PredictioninMedicine: StatisticalModelsversusArtifi- cialNeuralNetworks (cid:4) AnaPapoila,BioestatísticaeInformática,FaculdadeCiências MA-11 MédicasdaUniversidadeNovadeLisboa,CEAUL,Portugal, Monday,9:00-10:20 [email protected],CarlosGeraldes,PatriciaXufre 8.2.38 Artificial Neural Networks are often used in Biomedical Sciences and Emerging Data Mining Applications in Medicine. Amaingoalistopredictaclinicaloutcomeaftertakingintoac- countasetofindependentexplanatoryvariables.ANNsariseasanalternative Biomedics and Biotech to logistic regression. This study compare Generalized Linear Models with binaryresponse,withtheperformanceofANNs,inwhatconcernstheirpre- Stream: EmergingApplicationsofOR dictiveanddiscriminativepower. Forbothapproaches,validationtechniques Invitedsession wereapplied. Thesemethodologieswereusedtopredictmortalityofpatients admittedtoanIntensiveCareUnitlocatedinLisbon. Chair:HonoraSmith,SchoolofMathematics,Universityof Southampton,Highfield,SO171BJ,Southampton,Hampshire, UnitedKingdom,[email protected] 5- A Study in Different Channels’ Consumer on the Pur- Chair:Gerhard-WilhelmWeber,InstituteofAppliedMathematics, chasingIntentionandBehaviorofBio-technologyProd- MiddleEastTechnicalUniversity,ODTÜ,06531,Ankara,Turkey, ucts [email protected] YuanchauLiour,LogisticsManagement,TakmingUniversityof Chair:VeronicaBiga,DepartmentofAutomaticControlandSystems ScienceandTechnology,11451,Taipei,Taiwan, Engineering,TheUniversityofSheffield,MappinStreet,S13JD, [email protected],Chiao-LingHuang,Chie-beinChen Sheffield,Afghanistan,v.biga@sheffield.ac.uk Biotechnologyhasbeenplayinganimportantroleinmodernfinancialsoci- 1- The Information Effect of the Infectious Diseases Out- ety;recently,asimprovementofeconomyenvironmentandthedevelopment breakonBiotechnologyStockPerformance oftechnology,thegovernmentiscommittedtotheimplementationofbiotech- nology. Weexploretheimpactofconsumerattitude,consumer’spurchasing Yi-HsienWang,DepartmentofBanking&Finance,Chinese intention,promotionsandproductinvolvement,andperceiveriskonpurchas- CultureUniversity,55,Hwa-KangRoad,Yang-Ming-Shan., ingintention. WetakeNorthernTaiwanarea’sconsumerastheresearchob- Taipei,Taiwan11114,R.O.C,11114,Taipei,Taiwan, jects,useSEMtoanalyzeandmakesuggestionstobio-technologyhealthfood [email protected],Fu-JuYang,Kuang-HusnShih, companiesinaccordancewiththeempiricalconclusions. Li-JeChen 5 MA-12 EURO24-Lisbon2010 (cid:4) MA-12 Monday,9:00-10:20 8.2.39 (cid:4) MA-13 AHP 01 Monday,9:00-10:20 2.2.21 Stream: AnalyticHierarchyProcesses,AnalyticNetwork Processes Location and GIS Invitedsession Stream: LocationAnalysis Chair:GrzegorzGinda,Dept.ofOperationalResearchin Invitedsession Management,OpoleUniversityofTechnology,Facultyof ManagementandProductionEngineering,ul.Warynskiego4, Chair:IoannisGiannikos,BusinessAdministration,Universityof 45-057,Opole,Poland,[email protected] Patras,UniversityofPatras,GR-26500,Patras,Greece, [email protected] 1- AviationandtheBelgianClimatePolicy(ABC)-AMul- ticriteriaAnalysis(MCA)fortheevaluationofpolicyop- 1- Enhancing Location Optimization Modeling Capabili- tions to mitigate the total aviation climate change im- tiesthroughtheuseofGIS pact AlanMurray,GeographicalSciencesandUrbanPlanning, AnnaliaBernardini,MOSI-T,VrijeUniversiteitBrussel, ArizonaStateUniversity,P.O.Box875302,85287-5302,Tempe, Pleinlaan2,1050,Brussels,VlaamseBrabant,Belgium, AZ,UnitedStates,[email protected] [email protected],TomVanLier,Annelies Theprevalenceofwidelyavailableandaccessiblegeographicinformationsys- Heemeryck,EllenVanHoeck,CathyMacharis tem(GIS)packagesandassociatedgeographicdatahasbeenimportanttoall TheABCprojectanalysesthedifferentclimatepolicyoptionsaimedtoreduce disciplinesthatstudy,analyzeandevaluatespatialproblems. Inthispaperwe theclimatechangeimpactsoftheaviationsector. Inviewtocomparethedif- summarizemajorcharacteristicsofGISrelevanttolocationmodelingandspa- ferentalternativepolicies(finan.&econ. tools,R&D,operat. procedures)the tialoptimization. Anumberofwidelyrelieduponoptimizationmodelsare MCAmethodisapplied. Theperformancesofthosepoliciesareevaluatedin detailed. Particularattentionisgiventoidentifyingimplementationandap- relationtosomeappropriatecriteria(env. performances,social-economicim- plicationlimitations,andhowthesecanbeovercomethroughintegrationwith pactsaviationsector).AcombinationoftheAnalyticHierarchyProcessandthe GIS. PROMETHEEmethodallowstocometoadetailedanalysisoftheadvantages anddisadvantagesofeachoftheproposedpolicymeasures. 2- The Geographical Information System ’Ptolemeos- Europe’andtheanalysisofregionalgeo-economicdy- 2- A consistency-based method for aggregating prefer- namicsofFrance ence information from multiple pairwise comparison matrices JohnKarkazis,BusinessSchool,UniversityoftheAegean, Chios,GR-82100,Chios,Greece,[email protected] EstherDopazo,LenguajesySistemasInformáticos,Universidad PolitecnicaMadrid,FacultadInformática,Campusde Thispaperexplorestheregionalgeo-economicdynamicsofFranceduringthe period1985-2004employingGIS"Ptolemeos-Europe’. Inthebeginning,the Montegancedo,28660,BoadilladelMonte,Madrid,Spain, keynotionofregionalefficiencyisintroducedaswellasotherkeyregional edopazo@fi.upm.es,MauricioRuiz-Tagle analysisnotionssuchas: regionaldiscriminationcostandregionaldiscrimi- Weconsideragroupdecisionproblem,wheredecisionmakersestimatetheir nationiso-curves. Basedonthenotionofregionalefficiencythegeneralgeo- preferencesofasetofalternativesintotheformofpairwisecomparisonmatri- economicgravitymodelisintroducedanditsoutcome,thegeo-economicgrav- ces(awell-establishedtechniqueinthisfield). Inthisscenario,afundamen- itycentersandotherrelatedstrategicanddynamicnotionssuchasthecapital talproblemisthegenerationofapriorityvectorforthealternativesfromthe displacementfactorandthevelocityofgeo-economicgravitycentersarepre- pairwisematriceswhichrepresentstheconsensusopinionforthegroup. We sentedanddiscussed. TheaboveareappliedtoFranceinordertoexplorethe proposeaweightedlogarithmicgoalprogrammingmethodforaggregatingin- strategicgeo-economictrendsofthe22administrativeregionsofit. dividualopinionsintoanoptimalgrouppriorityvector,wheretheconsistency ofeachexpertistakenintoconsideration. 3- Improving the efficiency of WEEE collection systems usingaweb-basedGISapplication 3- An agreement-based approach for generating priority vectorsfrommultiplepairwisecomparisonmatrices SimãoRibeiro,ProductionandSystemsDepartment,University ofMinho,Portugal,UniversidadedoMinho,CampusdeGualtar, MauricioRuiz-Tagle,FacultaddeCs.delaIngeniería, 4710-057,Braga,Portugal,[email protected],Jorge UniversidadAustraldeChile,GeneralLagos2086,Campus Pereira,JoelCarvalho,JoséOliveira,ManuelFigueiredo,José Miraflores,Valdivia,Chile,[email protected],EstherDopazo Telhada,LuisDias Theproblemofimportanceweightsanalysisanddeterminationfrommultiple source information is a critical issue in many fields such as machine learn- Thisprojectfocusesonthedesignofaweb-basedGISapplicationtosupport ing,meta-searchengines,multi-criteriadecisionmaking,etc.Wefocusonthe theplanningandmanagementofcollectingwasteofelectricalandelectronic problemofcomputingtheimportanceweightsandthecorrespondingrankor- equipments(WEEE)networks.Itaddressestheissuesofgatheringandmanag- deringofasetofalternativesfrominformationgivenbyagroupofexperts inginformationneededbynetworkoptimizationmodulesbeingincludedinan intotheformofpairwisecomparisonmatrices.Wepresentanapproachbased integratedcomputerizedapplication,andtheissuesofanalyzingandmapping onlpdistance-basedaggregationfunctionsandontheuseofconsensus-driven theiroutputs. SeveralGISanddatabasetechnologieswillbeused,andtheir weightsforquantifyingtherelativeimportanceoftheexperts. applicabilityandutilitywillbediscussed.Inoverall,itisexpectedthatrelevant economicandenvironmentalbenefitswillbeachieved. 4- IntegratedMADAAssessmentTool 4- MultiobjectiveDemandCoveringModelsbasedonGIS GrzegorzGinda,Dept.ofOperationalResearchinManagement, OpoleUniversityofTechnology,FacultyofManagementand IoannisGiannikos,BusinessAdministration,Universityof ProductionEngineering,ul.Warynskiego4,45-057,Opole, Patras,UniversityofPatras,GR-26500,Patras,Greece, Poland,[email protected],MiroslawDytczak [email protected],GeorgiosAlexandris Integratedtoolforinterdisciplinaryassessmentofdecisionalternativesinman- Inthispaperwediscussanumberofmaximaldemandcoveringmodelswhere agementandengineeringisdiscussedinthepaper.Thetoolmakesuseofsev- the customers as well as the servers may be geographic objects rather than eralselectedMADAapproachestoobtainmorediversifiedresults.Acommon singlepoints. ThroughtheuseofGeographicInformationSystems(GIS),we datastructureisappliedtomakepreparationofrequireddatalessexpensivein considerdifferentnotionsofcoverageanddevelopaseriesofmultiobjective termsofbothtimeandworkeffort. Componentmethodsmakeitpossibleto programmingmodelsthattakeintoaccountthefollowingobjectives:(a)max- includeandassessinfluenceofbothintangibleandtangibleaspects. Theyare imizationoftotalcoverage, (b)maximizationofminimumcoverageand(c) alsoeasilyimplementable.Thetooladdressesissuesofinputdataconsistency minimizationofdistancetoserversofuncovereddemandobjects.Thesemod- andgroupdecisionsupporttoo. Thetoolisuniquewithregardtoabilityof elstakeintoaccountthegeographyofeachdemandareainquestionandadjust adaptationtoparticularneeds. Sampleanalysisisincludedwhichshowsits thelocationoftheserversaccordingly. applicabilityandscaleofpotentialapplicationbenefits. 6 EURO24-Lisbon2010 MA-15 (cid:4) (cid:4) MA-14 MA-15 Monday,9:00-10:20 Monday,9:00-10:20 2.2.15 2.2.12 Inventories in Supply Chains Location-routing problems Stream: SupplyChainPlanning Stream: VehicleRouting Invitedsession Invitedsession Chair:HorstTempelmeier,SupplyChainManagementand Chair:ChristianPrins,LaboratoireLOSI,UniversitédeTechnologie Production,UniversityofCologne,Albertus-Magnus-Platz,D-50923, deTroyes,BP2060,10010,TroyesCedex,France,[email protected] Cologne,Germany,[email protected] 1- Centralized Distribution System of Infusion Solutions 1- A discrete time multi-level inventory system with a onaNetworkofHealthCareUnits make-to-ordersupplier WilliamGuerrero,IndustrialEngineering,Universidaddelos HorstTempelmeier,SupplyChainManagementandProduction, Andes,AvCll147No17-81ap502,472,bogotá,bogotáD.C., UniversityofCologne,Albertus-Magnus-Platz,D-50923, Colombia,[email protected],NubiaVelasco,Ciro Cologne,Germany,[email protected] AlbertoAmaya,ChristelleGueret,ThomasYeung We study asupply network comprising afactory following a make-to-order Amethodologytoimproveinventorycontrolanddistributionpoliciesinhos- strategy,awarehouseusingareorderpoint-reorderquantitypolicyanddistribu- pitalsispresented. Thestrategyistocentralizethemanagementofmedicines tioncentersusingbase-stockpoliciesindiscretetime. Thefactoryismodeled intoasingledepottoreducecosts.Itisaimedtofindoptimalinventorycontrol asadiscretetimeG/G/1queueingsystem.Thesystemisdecomposedintothree policiesforone-warehousen-retailerdistributionsystembasedonaMarkov layersthatarelinkedthroughrandomwaitingtimes. Anoveralloptimization Chainmodel.Resultsareevaluatedonarealhospital.AnMIPmodeltodecide modelincludingasdecisionvariablestheprocessingtimeinthefactoryand thelocationofthecentraldepotanddistributionroutestotheCareunitsisalso theparametersoftheinventorypoliciesappliedinthedistributionsystemis proposed.Theobjectiveistheminimizationofthecostsoftheprojectandthe formulatedandsolved. inventory-on-handvalueofthesystem. 2- Two-Capacitated-Supplier Two-Stage Periodic-Review 2- Solution methods for the periodic location-routing SupplyChainProblemInvestigation problem KaiLuo,OperationsManagement&InformationTechnology, CarolineProdhon,UniversityofTechnologyofTroyes,12rue HECParis,1,ruedelaLibération,78351,JouyenJosas,Paris, MarieCurie,10000,Troyes,France,[email protected] France,[email protected],LaoucineKerbache,Ramesh Thewell-knownVehicleRoutingProblem(VRP)hasbeendeeplystudiedover Bollapragada thelastdecades. Nowadays,generalizationsaredevelopedtowardtacticalor strategicdecisionlevelsbutnotboth. ThetacticalextensionorPeriodicVRP Inthispaper,weinvestigateatwo-product(high-end,low-end),oneretailer/ (PVRP)plansasetoftripsoveramultiperiodhorizon.Thestrategicextension twosuppliersproblemwithrandomdemandandperiodicreview. Theprob- orLocation-RoutingProblem(LRP)ismotivatedbyinterdependentdepotloca- lemisdecomposedintosub-modelsandsolvedsequentially.Forsimplecases, tionandroutingdecisions.Thegoalhereistopresenttheveryrecentmethods closedformexpressionsareprovidedfortheoptimalsolution. Weshowthat, thatsolvethePeriodicLRP,acombinationofthePVRPandLRPintoaneven undercertainconditions,theretailershouldplacethehigh-endproductinthe morerealisticproblemcoveringalldecisionlevels. secondaryinventory. Forcomplexcases,weproposeaheuristictosolvethe problemandprovidemanagerialinsights. 3- Acuttingplaneapproachforthesingletruckandtrailer routingproblemwithsatellitedepots(STTRPSD) 3- Heuristicsformulti-item,two-echeloninventorysystem withaggregatemeanwaittimeconstraint. JuanG.Villegas,LOSI,UniversitedeTechnologiede Troyes/UniversidaddelosAndes,12,rueMarieCurie,BP2060, ArjunSubramaniam,AppliedMaterials,MountainView, 10010,Troyes,France,[email protected],JoseM. California,UnitedStates,[email protected], Belenguer,EnriqueBenavent,AntonioMartinezSykora, DeepakBhatia ChristianPrins,CarolineProdhon Weconsideramulti-item,two-echelonsparepartsinventorysystem,withone IntheSTTRPSDatruckwitharemovabletrailerbasedatamaindepotserves centralwarehouseandmultiplelocalwarehouses.Wepresentclose-to-optimal, thedemandofasetofcustomersreachableonlybytruck.Thus,beforeserving scalableheuristicstominimizetotalcostwitheachlocalwarehousesubjectto thecustomersintruckroutes,itisnecessarytodetachthetraileratappropriate anaggregatemeanwaittimeconstraint.Alllocationsoperateunderacontinu- parkingplacesandtotransferproductsbetweenthetruckandthetrailer. We ousreviewsystemwithbasestockpolicies.Wetesteffectivenessbycomparing presentatwoindexformulationoftheSTTRPSDandvalidinequalitiesthat withthelowerboundanddemonstratebetterperformancecomparedtoresults areusedwithinacuttingplanemethodtoproducelowerbounds,andtosolve fromrecentlypublishedworks. theproblemwithbranchandcut.Theresultsarecomparedwithupperbounds foundbyGRASP/VNDandILSon32randominstances 4- Price and Perception - Understanding the Consumer SideofRecoveredProducts 4- AhybridGRASPxPathRelinkingfortheTwo-Echelon LocationRoutingProblem JonathanLinton,SchoolofManagement,UniversityofOttawa, 39SachsForestPlace,K2G6V2,Ottawa,Ontario,Canada, VietPhuongNguyen,LOSI-UniversitédeTechnologyde [email protected],LeilaHamzaoui Troyes,Troyes,France,[email protected],Christian Prins,CarolineProdhon Weseektoaddressthegapinunderstandingconsumerwillingnesstopayfor productsthatarecomprisedofrecoveredmaterialsandparts. Consequently,a ThispaperpresentsahybridbetweenGRASPandPathRelinkingtosolvethe surveyof320respondentswasconductedtodeterminethewillingnesstopay Two-EchelonLocationRoutingProblem(LRP-2E).TheGRASPreinforcedby fordifferenttypesofproductscontainingrecoveredmaterialsandcomponents. aLearningProcessusesthreeconstructiveheuristicstogeneratetheinitialsolu- Aseriesofrelatedhypothesesareprovidedandtested.Inadditiontoconsider- tions.ThePath-relinkingaddsamemorymechanismbycombiningintensifica- ingissuespersonalattitudestotheenvironmentandperceivedrisk,weconsider tionstrategyandpost-optimization.Ourmethoduseslocalsearchesstructured theeffectsofbrandingandproductcharacteristics.Whiletheworkisempirical byaVariableNeighbourhoodDescent(VND).Computationalresultsconfirm innatureitiscriticaltosupplychainplanningasthereislimitedresearchand theefficiencyofthisapproachontwosetsofLRP-2Einstances. Furthermore understandingoftheconsumersideofclose-loopedsupplychains. itiscompetitivewiththebestmeta-heuristicpublishedfortheLRP. 7 MA-16 EURO24-Lisbon2010 (cid:4) (cid:4) MA-16 MA-17 Monday,9:00-10:20 Monday,9:00-10:20 2.2.14 1.3.14 Rolling stock and Re-scheduling Collaborative Planning I Stream: PublicTransport Stream: TransportationPlanning Invitedsession Invitedsession Chair:MarkusReuther,Optimization,Zuse-InstitutBerlin, Chair:HerbertKopfer,DepartmentofBusinessStudies& Takustrasse7,14195,Berlin,Germany,[email protected] Economics,ChairofLogistics,UniversityofBremen, Wilhelm-Herbst-Strasse5,28359,Bremen,Germany, 1- RollingStockRotationPlanningforIntercityRailTraffic [email protected] Chair:MelanieBloos,ChairofLogistics,BremenUniversity, MarkusReuther,Optimization,Zuse-InstitutBerlin,Takustrasse WilhelmHerbstStr.5,28359,Bremen,Germany, 7,14195,Berlin,Germany,[email protected] [email protected] 1- Transportation Operations Planning and Cost Alloca- Weconsideroneofthebasicoperationalplanningproblemsinpublicrailtrans- port,theconstructionofarollingstockschedule. Theproblemdealswiththe tioninaCooperativeScenario optimizationoffeasiblerotationsforindividualrailcarsand,simultaneously, AndreaNagel,Dept.ofInformationSystems,FernUniversität- thecompositionoftrainsetsfromtheserailcars. Inaddition,wehavetointe- UniversityofHagen,58084,Hagen,Germany, grateserveralmaintenanceandregularityaspects.Modelingandcomputational preliminiaryresultsforinstancesofourindustrialpartner,DBFernverkehrAG, [email protected],GiselherPankratz,Hermann whichoperatesapproximately1.300trainsinEuropeperday,arepresented. Gehring Cooperativescenariosintransportationplanningusuallyhavetocopewiththe taskofsolvinganoptimizationproblem,aswellasfindingafairallocationof 2- Railway Rolling Stock Rescheduling with Rerouting of thecostsamongthepartners.Weidentifyandcharacterizetheseproblemsfora Passengers real-lifecooperationoffourproducersinthefoodandbeveragesindustry,who decidedtocoordinatetheirdistributionactivitiesbyinter-organisationaltrans- GaborMaroti,DepartmentofDecisionandInformation portationplanning. Furthermore,wepresentasolutionmethodthathasbeen implemented,integratingaGRASPheuristicwiththeShapleyvalueapproach. Sciences,RotterdamSchoolofManagement,Erasmus Finally,weshowcomputationalresults. UniversityRotterdam,BurgOudlaan50,3062PARotterdam, TheNetherland,3062PA,Rotterdam,Netherlands, 2- Allocating Cost of Service to Customers in Inventory [email protected],LarsNielsen,LeoKroon Routing Inthispresentationwedescribedisruptionmanagementprocessesforapas- OkanOzener,IndustrialEngineering,OzyeginUniversity, sengerrailwaysystem.Inadisruptedsituation,thetimetable,therollingstock KusbakisiCadNo:2,AltunizadeUskudar,34662,Istanbul, circulation,andthecrewdutiesmustberescheduled. Wefocusonreschedul- Turkey,[email protected],OzlemErgun,Martin ingtherollingstockcirculation.Incaseofadisruption,thepassengersmaybe Savelsbergh willingtotakeadetourroutearoundthedisruptedarea.Thentherollingstock circulationmustberescheduledinsuchawaythatadditionalseatingcapacity Vendormanagedinventoryreplenishmentisacollaborationbetweenasupplier isprovidedalongthedetourroute. Inthispresentationwedescribeanitera- anditscustomerswherethesupplierisresponsibleformanagingthecustomers’ tiveprocedurethatreroutesthepassengers,andthatmodifiestherollingstock inventorylevels. InourVMIsetting,thesupplierexploitssynergiesbetween circulationaccordingly. Computationalresultsbasedonreal-lifeinstancesof customers,e.g.,theirlocations,usagerates,andstoragecapacities,toreduce NetherlandsRailwayshaveshownthatthisproceduremaysubstantiallyreduce distribution costs. Due to the intricate interactions between customers, cal- theirdelays. culatingafaircost-to-serveforeachcustomerisadauntingtask. However, cost-to-serveinformationisusefulwhenmarketingtonewcustomers,orwhen revisitingroutinganddeliveryquantitydecisions. Wedesignmechanismsfor thiscostallocationproblemanddeterminetheircharacteristicsbothanalyti- 3- RapidTransitNetworks: TimeTableandRollingStock callyandcomputationally. ÁngelMarín,MatemáticaAplicadayEstadística,Universidad 3- Collaborative vehicle routing in a multi-depot environ- PolitécnicadeMadrid,E.T.S.IngenierosAeronáuticos,Plaza ment CardenalCisneros,3,28040,Madrid,Madrid,Spain, JuliaRieck,DepartmentforOperationsResearch,Clausthal [email protected],LuisCadarso UniversityofTechnology,Julius-Albert-Str.2,38678, Clausthal-Zellerfeld,Germany,[email protected] Inrapidtransitnetworks, thedailyoperationsmanagementprocessincludes two major tasks: 1. Train services Timetable (TT). 2. Rolling Stock (RS) Fiercecompetitionurgescarrierstocooperate.Particularly,medium-sizedcar- assignmenttotheTT.Thetasksareinterdependentbutareoftensolvedse- riers only achieve the adequate area coverage by splitting transportation re- quentiallyduetorestrictionsoncomputationaltimeandtheintractabilityofan questsintomultipletasks(pick-up,linehaul,delivery)thatcanbehandledsep- integratedapproach. Inourmodelingapproachweconsidertheintegrationof aratelybydifferentcarriers. Hence,acarrierhastoperformthedeliveryand TTandRS.Somecomputationalexperimentswillbepresented. pick-upservicesaroundthedepotwhileminimizingthetransportationcosts.In ordertoimprovetheresultingsetofsingle-depotsolutions,wepresentanew collaborativemethodthattriestofindareassignmentoftaskstocarrierswhich 4- Assignmentofservicesinbuslinesundercongestion decreasestheoveralltransportationcosts. 4- Theevaluationofpickupanddeliveryrequestsincases EsteveCodina,StatisticsandOperationalResearch,UPC,Edifici C5,Desp216CampusNord,08034,Barcelona,Spain, ofasymmetricinformation [email protected],ÁngelMarín,FranciscoLopez MelanieBloos,ChairofLogistics,BremenUniversity,Wilhelm HerbstStr.5,28359,Bremen,Germany,[email protected], Amodelispresentedfordimensioningthenumberofservicesinbuslinesop- HerbertKopfer eratingundercongestionedsituations,whichmayariseincaseofdisruptionof Collaborativetransportplanningaimsatcreatingthemostefficientallocation aRapidTransitNetwork. Themodeltakesintoaccountbuscapacitylimita- ofrequeststocarriersforagroupagesystem. However,duetothenatureof tionsandfleetavailabilityaswellasthedwelltimesofbusesatstations.Also, thissystem,onlylimitedrelevantinformationonthecarriers’currentplanning ananalysisofthewaitingtimeofpassengersatbusstopsismadewithspecial isavailablesystem-wide.Ourresearchfocusesonevaluationcriteriathatcreate emphasisonthisfactoronthemodelresults.Themodelisformulatedundera anefficientsolutiondespiterestrictedinformationonthecarriers’situationand system-optimumpointofviewandaheuristicalgorithmapproachisdeveloped wepresentinitialresultsontheperformanceofevaluationcriteriaforindividual forlargersizenetworks. requests. 8 EURO24-Lisbon2010 MA-19 (cid:4) (cid:4) MA-18 MA-19 Monday,9:00-10:20 Monday,9:00-10:20 1.3.15 1.3.20 New Achievements in Stochastic Models Game Theory and Economics and Optimization Stream: DynamicalSystemsandGameTheory Stream: StochasticModelingandSimulation Invitedsession Invitedsession Chair:Gerhard-WilhelmWeber,InstituteofAppliedMathematics, Chair:ErikKropat,DepartmentofComputerScience,Universitätder MiddleEastTechnicalUniversity,ODTÜ,06531,Ankara,Turkey, BundeswehrMünchen,Werner-Heisenberg-Weg39,85577, [email protected] Neubiberg,Germany,[email protected] Chair:AlbertoA.Pinto,DepartamentodeMatematica,Universityof Minho,EscoladeCiências,UniversidadedoMinho,4710-057, 1- InflationDerivatives: HJMFrameworkandMarketMod- Braga,Portugal,[email protected] els 1- Investments under Oligopolistic Competition in a vin- Kwai-sunLeung,SystemsEngineeringandEngineering tagedifferentialgame Management,TheChineseUniversityofHongKong,Shatin, StefanWrzaczek,UniversityofTechnologyVienna,Instituteof NewTerritories„HongKong,N.A.,HongKong,HongKong, MathematicalMethodsinEconomics,Argentinierstr.8,1.Stock [email protected],LixinWu (ORDYS),1040,Vienna,[email protected], PeterM.Kort Inthispaper,weestablishaHeath-Jarrow-Morton(HJM)typeframeworkthat governstheco-evolutionofthetermstructureofbothnominalandinflation Duetotechnologicalprogressnewmachinesaremoreproductive. Thusitis rates. Pricingofinflationderivativesunderthisframeworkcanbecarriedout notonlyimportanttochosetheoptimalamountofinvestments,butalsotheop- similarly to that of nominal interest-rate derivatives under the classic HJM timalage.Westudyanoligopolyconsistingoftwobigfirmsconnectedbythe model. Based on the HJM framework, we further develop a market model pricetheygetfortheproducts.Inordertoincludethevintageeffectofinvest- withsimpleforwardinflationratesusingdisplaceddiffusionprocesses,which mentswehavetocombinedistributedoptimalcontroltheorywithdifferential resultsinclosed-formpricingforinflationcapletsandinflationswaptions.The gametheory. Forthetwoindependentvariables(ageandtime)theconceptof smilemodelcanalsobedevelopedbasedonthemarketmodel. timeconsistencyhastobeconsideredforbothdirections.Importantresultsfor bothdirectionscanbederivedandmotivated. 2- Onsomeantagonisticgamerelatedtomajorityvoting 2- Strategic Interaction in Macroeconomic Policies: An ApplicationofDifferentialGameTheory MichaelKhachay,UralBranchofRAS,InstituteofMathematics KrishnaKumar,ManagingDirector,SamkhyaAnalyticaIndia andMechanics,S.Kovalevskoy,16,620990,Ekaterinburg, PvtLtd,110SobhaOpal39thCross,4thT-BlockJayanagar, RussianFederation,[email protected] 560041,Bangalore,Karnataka,India,[email protected],Puja Simplemajorityvotingisaclassicalapproachtoaggregationofindividualde- Guha cisionssuggestedbyacommitteeofexperts. Inthispaper,stabilityofsucha Theworldeconomyisnowhighlyconnected,enablingdevelopmentofdevel- collectivedecision,s.t. exclusionofsomefixednumberofexperts,isinvesti- opedanddevelopingcountries. Butitisaccompaniedbyrisksofpropagating gated.LetsomegivenlistLofdecisionsbeacceptedbysomecommitteeofq adverseshocksfromonecountrytoanother. Isolatedmacroeconomicpolicies equivalentexperts,andletsomenumberk<qbefixed.Howsmallcanthecar- guidedearlierthroughcontroltheoreticmodelsneedtobemodifiedintomod- dinalityofthelargestsublistL’ofL,acceptablebyanyk-memberssubcommit- elsofpolicycoordination.Exploitingtheconceptsofnon-cooperativesolution tee,be?Tightlowerboundofthisquantitywillbepresented,it’sasymptotical throughNash-Equilibrium,andcooperativesolutionthroughNashprogramin behaviorandapplicationswillbediscussed. differentialgameswithquadraticpay-offsweintroducestrategicinteractionin macroeconomicpolicy. 3- Theeffectofcorrelationinmake-to-ordersystems 3- An optimal selection of watchman routes by search MichaelZazanis,Statistics,AthensUniversityofEconomicsand game Business,76Patissionstr.,10434,Athens,Greece, ShuheiMorita,Dep.ofComputerScience,NationalDefense [email protected] Academy,1-10-20,Hashirimizu,239-8686,Yokosukacity, KanagawaPrefecture,Japan,[email protected],Ryusuke Westudytheeffectofcorrelationintheprocessingtimesofmake-to-ordersys- Hohzaki,ToruKomiya,EmikoFukuda temsusingmarkovianqueueingmodelsandmatrix-geometrictechniques.Par- ticularemphasisisplacedonthecorrelatedprocesswithexponentialmarginals Thispaperdealswithanoptimalselectionofwatchmanroutesinafacility,such derived from the Kibble-Moran-Downton bivariate exponential distribution. asartgallery. Artgalleryproblemisafamousprobleminthefieldofcom- Bothexactresultsandlogarithmicasymptoticsarederivedforthesesystems. putationalgeometry, whichdecidesthestationarydeploymentofwatchmen. Assemble-to-ordersystemswithcorrelateddemandsarealsoanalyzed. Thispaperconsidersadynamicpatrolplanforthefacilityfromtheviewpoint ofOperationalResearch. First, weobtainanoptimalintrusionschedulefor theminimumdetectionprobabilityoftheintruderbydynamicprogramming. 4- Rollover Optimization under Uncertain Regulatory Ap- Secondly,weformulateanoptimalselectionproblemofwatchmanroutesand provalDateforProductswithBassDemandRate solveitbysearchtheoryandgametheory. HibaElKhoury,OperationsManagementandInformation 4- Nonparametricpredictionofbankloanrecoveries Technology,HEC-PARIS,HECParis,1ruedeLaliberation, JoãoBastos,TechnicalUniversityofLisbon,CEMAPRE-ISEG, 78350,JouyenJosas,France,[email protected], RuadoQuelhas6,1200-781,Lisboa,Portugal, ChristianVanDelft,LaoucineKerbache [email protected] Consideracompanythatplanstophase-outanexistingproductandphase-in WiththeadventoftheBaselIIAccord,bankingorganizationsareinvitedtoes- areplacementone. Ifproductionoftheexistingproductisstoppedearly,the timatecreditriskcapitalrequirementsusinganinternalratingsbasedapproach. firmwillloseprofitandgoodwill. Yet,ifproductionoftheexistingproduct Inordertobecompliantwiththisapproach,institutionsmustestimatetheex- isstoppedlate,thefirmwillexperienceanobsolescencecostfortheexisting pectedloss-given-default,thefractionofthecreditexposurethatislostifthe product. WestudyrolloverofproductsthatfollowaBassdemandrate,with borrowerdefaults.Thisstudyevaluatestheabilityofnonparametricregression thenewproductsubjecttoanuncertainregulatoryapprovaldate.Weminimize treemodelstoforecastbankloancreditlosses. Theperformanceofthemod- costsassociatedwiththerolloveranddetermineoptimalrolloverstrategiesfor elsisbenchmarkedagainstrecoveryestimatesgivenbyhistoricalaveragesand differentproductfamilies. parametricfractionalresponseregressions. 9 MA-20 EURO24-Lisbon2010 (cid:4) (cid:4) MA-20 MA-21 Monday,9:00-10:20 Monday,9:00-10:20 1.3.33A 6.2.47 Cutting and Packing 1 OR in Practice I Stream: CuttingandPacking Stream: SoftwareforOR/MS Invitedsession Invitedsession Chair:JoseFernandoOliveira,FacultyofEngineering/INESC Chair:AnaMoura,Economics,ManagementandIndustrial Porto,UniversidadedoPorto,RuaDr.RobertoFrias,4200-465, Engineering,UniversityofAveiro,CampusUniversitáriode Porto,Portugal,[email protected] Santiago,3810-193,Aveiro,Portugal,[email protected] 1- Radical-free phi-functions for 2D objects and their ap- 1- Optimizing Fire Station Locations for Istanbul plications MetropolitanMunicipality TatianaRomanova,DepartmentofMathematicalModelingand EmelAktas,IndustrialEngineeringDepartment,Istanbul OptimalDesign,InstituteforMechanicalEngineeringProblems TechnicalUniversity,ITUIsletmeFakultesi,Macka,34367, oftheNationalAcademyofSciencesofUkraine,2/10Pozharsky Istanbul,Turkey,[email protected],OzayOzaydin,SuleOnsel, St.,61046,Kharkov,Ukraine,[email protected],Nikolai BurcinBozkaya,FusunUlengin Chernov,YuriStoyan,A.Pankratov Istanbulisadenselypopulatedcitywith2000+yearsofculturalheritage. We provideamax-covertypemathematicalmodelwithlimitedbudgettohelpMu- Phi-functionsareusedtosolvepacking,cutting,andcoveringproblems. Our nicipalityauthoritiesdeterminenewfirestationlocationsinadditiontoexisting purposeistoconstructphi-functionsbyusingsimplemathematicalformulas ones.WesolvethismodeltooptimalityusingGAMSandincreaseexistingfire withoutradicals. Firstweintroduceaspecialclassofbasicobjectsandprove coveragefrom56%to86%. Wealsoconsiderunlimitedbudget, overallre- thatany2Dobjectwhosefrontierisformedbycirculararcsandlinesegments location,andincreasedemphasisoncity’sculturaltreasures. WeuseGISto mayberepresentedasaunionofbasicobjects. Thenwederiveacomplete processallgeographicalinputdata,calculatenetworkdistancesandcoverage classofradical-freephi-functionsforallpairsofbasicobjects.Lastlyweshow ratios,andvisualizevarioussolutionsofourmodel. howtoformphi-functionsformoregeneralobjects. Asoftwarepackageis developedbasedontheseresults.Someapplicationsaregiven. 2- Designofaspatiallyexplicitmodeltooptimizethese- lectionofreforestationprojectsandthesizingofade- 2- A constructive algorithm for leather nesting in the au- tentiondamtoreducepeakrunoffinasmallwatershed tomotiveindustry JochenBreschan,DepartmentofEnvironmentalSciences,ETH PedroBrás,UniversidadedoMinho,4710-057,Braga,Portugal, Zurich,CHNK73.1,Universitaetstr.22,8092,Zurich, [email protected],CláudioAlves,J.M.Valériode Switzerland,[email protected],HansRudorfHeinimann, Carvalho,TelmoPinto RichardChurch Weaddresstheleathernestingprobleminthecontextofanautomotivecom- We address the problem of reducing potential flood hazards within a small pany. Inthis2-dimensionalproblem,irregularshapes(carseatscomponents) Alpine watershed, using combinations of 1) spatially explicit reforestation havetobecutfromanaturalleatherhidewithholes,defectsandqualityzones. projects and 2) a detention dam at the outlet in order to reduce peak storm Weproposeasolutionalgorithmbasedonaconstructiveprocedure. Wedis- runoff. Theobjectiveistooptimizethelocationofreforestationprojectsand cussthedifferentaspectsofthisprocedureandexplainthestrategicoptionson thesizingofthedetentiondamtocost-effectivelymeettargetsofpeakrunoff whichitisbased. Wealsobrieflydescribetheno-fitpolygonmethodusedto reduction. WeshowhowtocastthisproblemasaMixedInteger-LinearPro- guaranteevalidplacementsintheleatherhides. Computationalresultsonreal grammingoptimizationmodelthatissolvedusingCPLEX.Finally,wepresent instancesarepresented. anapplicationofthismodeltotheVogelbachwatershed(CH). 3- A Hybrid Meta-heuristic Approach for Non-standard 3- A two-stage packing procedure for a Portuguese trad- PackingProblemswithAdditionalConditions ingcompany GiorgioFasano,SpaceInfrastructures&Transportation,Thales AnaMoura,Economics,ManagementandIndustrial AleniaSpaceItalia,Str.AnticadiCollegno253,10146,Turin, Engineering,UniversityofAveiro,CampusUniversitáriode Italy,[email protected] Santiago,3810-193,Aveiro,Portugal,[email protected], AndreasBortfeldt Thisworkfocusesontheorthogonalpackingoftetris-likeitemswithinanon- rectangulardomain(withforbiddenzones),inthepresenceofadditionalcondi- Thisworkreportsonthedevelopmentofaprototypicaldecisionsupportsys- tions,suchasbalancing. Theoverallproblemisformulatedintermsofmixed tem,calledPackingandRoutingOptimizer(PRO),whichisdevisedtosolve integerprogramming. Sincenontrivialcasesgiverisetoverylarge-scalein- severalpackingandroutingproblemsforaPortuguesecompany.Thedailydis- stances,ahybridmeta-heuristicapproachhasbeenadoptedtosolverecursively tributionprocessisanalyzedandthreedecisionproblemsregardingautomated theproblem.Itisbasedontheconceptofabstractconfiguration,derivingfrom decisionsupportaredetermined. Theperformanceofthesolutionapproaches therelativepositionofitems.Anextensionconsidersthe2Dcaseofpolygons isevaluatedbycomputationaltestsbasedonactualcompanydata.Thetestre- fromaglobaloptimizationpointofview. sultsshowthatthesystemisabletohelpimprovingthedailydecisionsandto strengthentheflexibilityinnegotiationswithcustomers. 4- Dualfeasiblefunctionsforvectorpackingproblems JürgenRietz,DepartamentodeProduçãoeSistemas,Centrode (cid:4) MA-22 InvestigaçãoAlgoritmi,UniversidadedoMinho,Campusde Gualtar,4715-082,Braga,Portugal,[email protected], Monday,9:00-10:20 CláudioAlves,J.M.ValériodeCarvalho,FrançoisClautiaux 3.1.10 Dual-feasible functions (DFFs) were successfully used to obtain fast valid Teaching OR/MS lower bounds for the one-dimensional cutting stock problem. To accelerate thecalculations,onlymaximal,especiallyextremalfunctionsshouldbeused. Stream: TeachingOR/MS Thisapproachworksforthevectorpackingproblemtoo,ifthedomainofthe Invitedsession DFFsisreplacedbyamore-dimensionalunitcube.Westatenecessaryandsuf- ficientconditionsforsuchfunctionstobemaximalrespectivelyextremaland Chair:SusanaColaco,DepartamentodeCiênciasMatemáticase presentsomenon-trivialexamples. NewDFFsfortheproblemarediscussed, Naturais,EscolaSueriordeEducação-Institutopolitécnicode andcomputationalresultsarereported. Santarém,CompexoAndaluz,Apartado131,2000,Santarem, Portugal,[email protected] 10

Description:
Mar 2, 2014 Sheu-hua Chen, Distribution Manegement, National Chin-Yi We consider one of the basic operational planning problems in public rail trans- port, the .. that any 2D object whose frontier is formed by circular arcs and line
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.