第三Decision_Analysis2运营管理笔记

©2008Prentice-Hall,Inc.Chapter3ToaccompanyQuantitativeAnalysisforManagement,TenthEdition,byRender,Stair,andHannaPowerPointslidescreatedbyJeffHeylDecisionAnalysis©2009Prentice-Hall,Inc.©2009Prentice-Hall,Inc.3–2LearningObjectives1.Listthestepsofthedecision-makingprocess2.Describethetypesofdecision-makingenvironments3.Makedecisionsunderuncertainty4.UseprobabilityvaluestomakedecisionsunderriskAftercompletingthischapter,studentswillbeableto:Aftercompletingthischapter,studentswillbeableto:©2009Prentice-Hall,Inc.3–3LearningObjectives5.Developaccurateandusefuldecisiontrees6.ReviseprobabilitiesusingBayesiananalysis7.Usecomputerstosolvebasicdecision-makingproblems8.UnderstandtheimportanceanduseofutilitytheoryindecisionmakingAftercompletingthischapter,studentswillbeableto:Aftercompletingthischapter,studentswillbeableto:©2009Prentice-Hall,Inc.3–4ChapterOutline3.1Introduction3.2TheSixStepsinDecisionMaking3.3TypesofDecision-MakingEnvironments3.4DecisionMakingunderUncertainty3.5DecisionMakingunderRisk3.6DecisionTrees3.7HowProbabilityValuesAreEstimatedbyBayesianAnalysis3.8UtilityTheory©2009Prentice-Hall,Inc.3–5Introduction
Whatisinvolvedinmakingagooddecision?
Decisiontheoryisananalyticandsystematicapproachtothestudyofdecisionmaking
Agooddecisionisonethatisbasedonlogic,considersallavailabledataandpossiblealternatives,andthequantitativeapproachdescribedhere©2009Prentice-Hall,Inc.3–6TheSixStepsinDecisionMaking1.Clearlydefinetheproblemathand2.Listthepossiblealternatives3.Identifythepossibleoutcomesorstatesofnature4.Listthepayofforprofitofeachcombinationofalternativesandoutcomes5.Selectoneofthemathematicaldecisiontheorymodels6.Applythemodelandmakeyourdecision©2009Prentice-Hall,Inc.3–7ThompsonLumberCompanyStep1Step1––Definetheproblem
Expandbymanufacturingandmarketinganewproduct,backyardstorageshedsStep2Step2––Listalternatives
Constructalargenewplant
Asmallplant
NoplantatallStep3Step3––Identifypossibleoutcomes
Themarketcouldbefavorableorunfavorable©2009Prentice-Hall,Inc.3–8ThompsonLumberCompanyStep4Step4––Listthepayoffs
Identifyconditionalvaluesconditionalvaluesfortheprofitsforlarge,small,andnoplantsforthetwopossiblemarketconditionsStep5Step5––Selectthedecisionmodel
DependsontheenvironmentandamountofriskanduncertaintyStep6Step6––Applythemodeltothedata
Solutionandanalysisusedtohelpthedecisionmaking©2009Prentice-Hall,Inc.3–9ThompsonLumberCompany00Donothing–20,000100,000Constructasmallplant–180,000200,000ConstructalargeplantUNFAVORABLEMARKET($)FAVORABLEMARKET($)ALTERNATIVESTATEOFNATURETable3.1©2009Prentice-Hall,Inc.3–10TypesofDecision-MakingEnvironmentsType1:Type1:Decisionmakingundercertainty
DecisionmakerknowswithcertaintyknowswithcertaintytheconsequencesofeveryalternativeordecisionchoiceType2:Type2:Decisionmakingunderuncertainty
ThedecisionmakerdoesnotknowdoesnotknowtheprobabilitiesofthevariousoutcomesType3:Type3:Decisionmakingunderrisk
Thedecisionmakerknowstheknowstheprobabilitiesprobabilitiesofthevariousoutcomes©2009Prentice-Hall,Inc.3–11DecisionMakingUnderUncertainty1.Maximax(optimistic)2.Maximin(pessimistic)3.Criterionofrealism(Hurwicz)4.Equallylikely(Laplace)5.MinimaxregretThereareseveralcriteriaformakingdecisionsunderuncertainty©2009Prentice-Hall,Inc.3–12MaximaxUsedtofindthealternativethatmaximizesthemaximumpayoff
Locatethemaximumpayoffforeachalternative
Selectthealternativewiththemaximumnumber000Donothing100,000–20,000100,000Constructasmallplant200,000–180,000200,000ConstructalargeplantMAXIMUMINAROW($)UNFAVORABLEMARKET($)FAVORABLEMARKET($)ALTERNATIVESTATEOFNATURETable3.2MaximaxMaximax©2009Prentice-Hall,Inc.3–13MaximinUsedtofindthealternativethatmaximizestheminimumpayoff
Locatetheminimumpayoffforeachalternative
Selectthealternativewiththemaximumnumber000Donothing–20,000–20,000100,000Constructasmallplant–180,000–180,000200,000ConstructalargeplantMINIMUMINAROW($)UNFAVORABLEMARKET($)FAVORABLEMARKET($)ALTERNATIVESTATEOFNATURETable3.3MaximinMaximin©2009Prentice-Hall,Inc.3–14CriterionofRealism(Hurwicz)Aweightedaverageweightedaveragecompromisebetweenoptimisticandpessimistic
Selectacoefficientofrealismαααα
Coefficientisbetween0and1
Avalueof1is100%optimistic
Computetheweightedaveragesforeachalternative
SelectthealternativewiththehighestvalueWeightedaverage=αααα(maximuminrow)+(1–αααα)(minimuminrow)©2009Prentice-Hall,Inc.3–15CriterionofRealism(Hurwicz)
Forthelargeplantalternativeusingαααα=0.8(0.8)(200,000)+(1–0.8)(–180,000)=124,000
Forthesmallplantalternativeusingαααα=0.8(0.8)(100,000)+(1–0.8)(–20,000)=76,000000Donothing76,000–20,000100,000Constructasmallplant124,000–180,000200,000ConstructalargeplantCRITERIONOFREALISM(αααα=0.8)$UNFAVORABLEMARKET($)FAVORABLEMARKET($)ALTERNATIVESTATEOFNATURETable3.4RealismRealism©2009Prentice-Hall,Inc.3–16EquallyLikely(Laplace)Considersallthepayoffsforeachalternative
Findtheaveragepayoffforeachalternative
Selectthealternativewiththehighestaverage000Donothing40,000–20,000100,000Constructasmallplant10,000–180,000200,000ConstructalargeplantROWAVERAGE($)UNFAVORABLEMARKET($)FAVORABLEMARKET($)ALTERNATIVESTAT