資源描述：
《multivariate probability distributions》由會員上傳分享，免費在線閱讀，更多相關內(nèi)容在工程資料-天天文庫。
1、Chapter5MultivariateProbabilityDistributionsPhysicianshavenothingtodowithwhatiscalledthelawoflargenumbers,alawwhich,accordingtoagreatmathematician’sexpressionisalwaystrueingeneralandfalseinparticular.ClaudeBernard,Frenchphysiologist(1813-1878)5.1OrientationWefrequentlyhavetheoccasiontostu
2、dythemutualinteractionsbetweentwoormorevariables.Howusefulareastudent’scollegegradepointaverage(GPA)andscoreonthelawschooladmissionstest(LSAT)aspredictorvariablesforastudent’sfutureperformanceinlawschool?Howdoesdiversi?cationofaninvestmentportfolioreducerisk,andhowdoesita?ectthereturns?Fo
3、rexample,whenoilpricessoarsodothestockpricesofoilcompanies,whilethoseofautomobilemanufacturersandairlinesfall.Modernportfoliotheorybeginswithananalysisofthejointvariationofthedi?erentassetclasses(suchasstocks,bonds,andmutualfunds)intheportfolio.Inenvironmentalstudiestheairqualityofametrop
4、olitanregionisdeterminedbysimultaneouslymeasuringseveralpollutantssuchasozone,nitrogenoxide,hydrocarbons,etc.Beforeundertakingaplantoreducethepollutionlevelstheenvironmentalscientistmuststudytheirinteractionssincereducingthelevelofjustonepollutantcouldactuallyincreasethecombinedharmfule?e
5、ctsoftheothers.Statisticiansusethejointdistributionfunctiontostudythejointvariationsamongthevariables.Webeginwiththediscretecaseandfollowwiththecontinuouscase,theonlydi?erencebetweenthemisthat,inthelattercase,weusesomebasictoolsfrommultivariatecalculus.OrganizationofChapter1.Section5.2:Th
6、eJointDistributionFunction:DiscreteRandomVariables2.Section5.2.1:IndependentRandomVariables3.Section5.3:TheMultinomialDistribution4.Section5.4:MeanandVarianceofaSumofRandomVariables5.Section5.4.1:TheLawofLargeNumbersforSumsofIndependentandIdenticallyDistributed(iid)RandomVariables6.Sectio
7、n5.4.2:TheCentralLimitTheorem7.Section5.5:WhyStockPriceshaveaLognormalDistribution:AnApplicationoftheCentralLimitTheorem205206IntroductiontoProbabilityandStatisticsforScience,Engineering,andFinance8.Section5.5.1:TheBinomialLatticeModelasanApproximationtoaContinuousT
