Poisson混雜模型的極大似然估計(jì).pdf

1、山東大學(xué)碩士學(xué)位論文Poisson混雜模型的極大似然估計(jì)姓名：李敏申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：概率論與數(shù)理統(tǒng)計(jì)指導(dǎo)教師：魏剛20090522VI第O章AbstractAbstractInthecontextofeconomicglobalization，thedegreeofbankingindstry7SopeningisalsograduallyincreasingWhenourBankingparticipatesintheinter

2、nationaIfinanceactivities，howtoguardagainstthefinancialriskalreadyhasbeentheimportantissueitfacingInalloftherisksthecommercialbankfacing，thecreditrisksthroughouttheentireoperatingprocess，whichhighlightstheimportanceofcre

3、ditriskmanagementInthisarticle，weintroducedthetwotraditionaImethodsofcreditriskmeasurementbrieflyaswellasfourmodernrisksmanagementmodels—CreditMetrics，KMVCPVandtheCreditRiskmodelThenwehavecomparedtheiradvantagesanddisadv

4、an—tagesThedeterminationofprobabilityofdefaultiSthecorecontentof”NewBaselCapitalAccord“InthearticleweintroducedthreecommonlyusedmethodsofcalculationofdefaultprobabilityAndthen，wefocusonLogisticModelwhichisthemostcommonly

5、usedinpracticeThefollowingiSPoissonmixedmodelmodelingofthedefaultintensitydirectlywhichistheinnovationofthearticleInthispaper，wehaveshowedthereasonablenessoftheproposalofthePoissonmixedmodel，thenestimatetheparametersusin

6、gthemaximumlikelihoodestimationmethodsIntheprocess，wehaveusedtherecursionformulaofnormaldistribution7SkmomentAsaresultofthecomplexityofthelikelihoodfunction，withtheanalysismethodestimateparameterisnotfeasibleFthereforeff

7、orestimatingtheparameterslweproposetousenumericalmethodstosolvethemaximumlikelihoodfunctionThenH—mericalmethodusedinthearticleiSthesteepestdescentmethodSupposethereisonlyonemacroeconomicindicatoraffectthedefaultintensity

8、thenmodelingbasedonthesupposeandtakeitasanexampleToestimateparametersofthePoisonmixedmodelusingthenumericalmethods孫eestimationoftheparam—etersandthenormaldistributionhaveagoodfitness，whichshowthatourmodeliSreasonableAndi