基于g函數(shù)的多元copula的構(gòu)造.pdf

1、西南交通大學(xué)碩士學(xué)位論文基于g函數(shù)的多元copula的構(gòu)造姓名：曾霞申請學(xué)位級別：碩士專業(yè)：應(yīng)用數(shù)學(xué)指導(dǎo)教師：何平20060301西南交通大學(xué)碩士學(xué)位論文第Ⅱ頁AbstractD印endencerelationsbc押c鋤珊d咖Variablesisoneofthcmostwidclystudiedsubjeccsinprobabili哆andstatisticsButmetraditionalcorrclationcoc伍cicⅡts

2、havesomelinlitsforame貓u億oft1IcdcpendencebeMeenr蛆domvariablcsUsingcopulaforme髂uringdependeⅡcctal【esmoreattentio璐inreccntyearsActIIally，c叩ulaplays趾iI】1ponantmleinthestudy0fdepcndencc，italsOcanbellscdforconstnlclingmultiVar

3、iatedistributionsⅡmcrecxistaf細(xì)ily0fcOpulas，wecouldconstnlctanybiVariateormunivariatedistfibutionfunctionsofgiveⅡmar舀nsbyS】(1ar’stheoremandthesedistributionsarealwaysveryusefIllforsimulationBecausethis，itismeaningflllof∞n

4、structingmultivariatecopulasButconstnlctingncopulasisdi衄cult舢tlloughthercexistmaIIywaysaboutconstnlctingbivariatecopulas，fcwofthcmc&nextendtomultivariate。SomemeIhodsabOutconstmcIingn—copulasinm鋤yrefcrenccshavelimitstosom

5、edegreeandalsocomplicatepremi∞Thispaper百vesanewmethodofconstnlctingncopulasdependsonthelawoftotalpTobabilityThemethodtllatc0船tnlctingn—copulasbascdong缸碰ionswhichde緬edinthispapertWecangctncopulasifaf缸myofgfullctionscouldb

6、cfoundFirsny’tllispaperin仃oducesthedefiIlitiOnaIIdbasicpropeniesofcopula，andsomercslIltsaboutoDnstmctionofncopulasinrecentyearsSecondly，thcmcthodofconstmctmg2copulaswasgiven，andtllenextendsmcconstnlctionprocedurcst0noopu

7、lasAndwehavediscIIssedsomepropcniesofthe2copulaswhich∞璐tructbythismcthodMcanwhileallpropcrtiesOf2一coplll弱iⅡthispaperhavendime璐ional柚alogsSinccthemethoddcpendsOngfIlnctio璐，丘ndinggfll刪onsisthekeytoconstnlctingncopulaS融ally

8、thispapcrhavcgiv％amethodoffindinggfunctiOnsd印ends0n0nedim肌sionaldistributionsAnd也crelatio璐be錒eeng西南交通大學(xué)碩士學(xué)位論文第Ⅱ頁AbstractD印endencerelationsbc押c鋤珊d咖Variablesisoneofthcmostwidclystudiedsubjeccsinprobabili哆andstatisticsButme

9、traditionalcorrclationcoc伍cicⅡtshavesomelinlitsforame貓u億oft1IcdcpendencebeMeenr蛆domvariablcsUsingcopulaforme髂uringdependeⅡcctal【esmoreattentio璐inreccntyearsActIIally，c叩ulaplays趾iI】1ponantmleinthestudy0fdepcndencc，italsOc

10、anbellscdforconstnlclingmultiVariatedistributionsⅡmcrecxistaf細(xì)ily0fcOpulas，wecouldconstnlctanybiVariateormunivariatedistfibutionfunctionsofgiveⅡmar舀nsbyS】(1ar’stheoremandthesedistributionsarealwaysveryusefIllforsimulatio

11、nBecausethis，itismeaningflllof∞nstructingmultivariatecopulasButconstnlctingncopulasisdi衄cult舢tlloughthercexistmaIIywaysaboutconstnlctingbivariatecopulas，fcwofthcmc&nextendtomultivariate。SomemeIhodsabOutconstmcIingn—copul

12、asinm鋤yrefcrenccshavelimitstosomedegreeandalsocomplicatepremi∞Thispaper百vesanewmethodofconstnlctingncopulasdependsonthelawoftotalpTobabilityThemethodtllatc0船tnlctingn—copulasbascdong缸碰ionswhichde緬edinthispapertWecangctnc

13、opulasifaf缸myofgfullctionscouldbcfoundFirsny’tllispaperin仃oducesthedefiIlitiOnaIIdbasicpropeniesofcopula，andsomercslIltsaboutoDnstmctionofncopulasinrecentyearsSecondly，thcmcthodofconstmctmg2copulaswasgiven，andtllenextend

14、smcconstnlctionprocedurcst0noopulasAndwehavediscIIssedsomepropcniesofthe2copulaswhich∞璐tructbythismcthodMcanwhileallpropcrtiesOf2一coplll弱iⅡthispaperhavendime璐ional柚alogsSinccthemethoddcpendsOngfIlnctio璐，丘ndinggfll刪onsist