11120.相依樣本下含附加信息時概率密度函數(shù)的經(jīng)驗(yàn)似然置信區(qū)間

1、相依樣本下含附加信息時概率密度函數(shù)的經(jīng)驗(yàn)似然置信區(qū)間EmpiricalLikelihoodConfidenceIntervalsforProbabilityDensityFunctionsinPresenceofAuxiliaryInformationUnderAssociatedSamples作者：董凱Author：塹壘i里Q壘g指導(dǎo)教師：秦永松專業(yè)：統(tǒng)計學(xué)學(xué)位：理學(xué)碩士授予單位：浙江師范大學(xué)Supervisor：!魚望g璺Q旦gQi

2、翌Major：曼圭璺主重量垡c璺Degree：叢壘鞋叢ofScienceInstitute：——ZhejiangNo——rmal——Un——iv——ers———i—ty——May，2016AbstractInthisarticle，wemainlyinvestigatetheproblemoftheempiricallikelihood(EL)confidenceintervalsforprobabilitydensityfunctio

3、nsinpresenceofauxiliaryinformationundernegativelyassociated(NA)samplesandstrongmixingsamplesUnderauxiliaryinformationconditions，theasymptoticdistributionoftheempiricallikelihoodratiostatisticforprobabilitydensityfunction

4、sisobtainedThisarticleincludesthreechapters：ThefirstchapterisintroductionTheconceptsandresearchprogressofNAsamplesandstrongmixingsamplesarereviewed，andtheresearchprogressofELmethodisalseviewedWealsodescribethestructurean

5、dcontentsofthisthesisinthischapterThesecondchapterstudiestheELconfidenceintervalsforprobabilitydensityfunc—tionsinpresenceofauxiliaryinformationundernegativelyassociated(NA)samplesWepresentthespecificformofthelikelihoodr

6、atiostatisticbyusingtheblockwisetech—niqueanddiscussitsasymptoticdistributionTheasymptoticdistributionoftheELratiostatisticforprobabilitydensityfunctionsinpresenceofauxiliaryinformationisprovedtobex2distributionfinallyTh

7、eresultisusedtoobtainELbasedconfidenceintervalsontheprobabilitydensityThethirdchapterstudiestheELconfidenceintervalsforprobabilitydensityfunctionsinpresenceofauxiliaryinformationunderstrongmixingsamplesSimilarlytothecase

8、ofNAsamples，wehaveprovedthattheasymptoticdistributionoftheELratiostatisticforprobabilitydensityfunctionsinpresenceofauxiliaryinformationunderstrongmixingsamplesisX2distributionTheresultisusedtoobtainELbasedconfidenceinte

9、rvalsontheprobabilitydensityTheinnovationofthisthesisismainlyreflectedinthefollowingthreeaspects：1WeobtaintheblockELratiostatisticinpresenceofauxiliaryinformationbyusingtheblockELmethod，andtheasymptoticdistributionoftheb