31303.birnbaumsaunders混合效應模型在定時截尾下的bayes推斷

1、扉頁：獨創(chuàng)性聲明本人聲明所呈交的論文是我個人在導師指導下進行的研究工作及取得的研究成果。除了文中特別加以標注和致謝的地方外，論文中不包含其他人或集體已經(jīng)發(fā)表或撰寫過的研究成果，對本文的研究做出貢獻的集體和個人均已在論文中作了明確的說明并表示了謝意。研究生簽名：雄日期：二翌啦論文使用和授權說明●本人完全了解云南大學有關保留、使用學位論文的規(guī)定，即：學校有權保留并向國家有關部門或機構送交學位論文和論文電子版；允許論文被查閱或借閱；學?？梢怨?/p>

2、布論文的全部或部分內容，可以采用影印、縮印或其他復制手段保存論文。(保密的論文在解密后應遵循此規(guī)定)研究生簽名：五婢導師本人及導師同意將學位論文提交至清華大學“中國學術期刊(光盤版)電子雜志社”進行電子和網(wǎng)絡出版，并編入CNKI系列數(shù)據(jù)庫，傳播本學位論文的全部或部分內容，同意按《中國優(yōu)秀博碩士學位論文全文數(shù)據(jù)庫出版章程》規(guī)定享受相關權益。研究生簽名：導師簽名：日期：AbstractBirnbaumSaundersdistribution

3、iSoneofthemostimportantdistributionsinreliabilityengineeringstudiesItisausefulmodeltodescribefatigueandreliabilitydataItdescribesthetotaltimeuntilthefatiguetosometypeofcumulativedamageBimbaumSaundersprobabilitydensityfun

4、ctioniscomplex，itisquitedifficulttoestimateurlknownparametersTherefore，toexploretheeffectiveparameterestimationmethodofBirnbaumSaundersdistributionhasgreatsignificanceandapplicationvalue，Therea他lotsofclustereddataorcorre

5、lateddatainthefieldsofengineering，biomedicine，economics，sociologyamongothersUsuallywea塢interestednotonlyinthegeneralpopulationsometimeswea托alsointerestedindifferencesbetweenindividualsMLxedeffectsmodelprovidesaneffective

6、methodtosolvethedatacorrelationheterogeneityoverdispcrsionBayesianmethodcombineswiththepriordistributionoftheparametersandsampleinformationwewillbeabletogetbetterstatisticalinferen∞Themajorityofmodelsbased011theBirnbaumS

7、aundersdistributionhaveassumedfixedeffects，andafewhavebeeninvestedforcorrelateddataThepurposeofthispaperistodevelopaBayesianstatisticalinferenceforBirnbaumSaundersmixedmodelsforcensoreddataUsually，itisverydifficulttoobta

8、intheparameterestimationin‘theprocessoftheBayesianmethodwhichinvolvesthecomplexandhigherdimensionalinte倒。塢SOthispaperusestheMCMCmethodtoestimateAhybridisobtainedthroughtreatingrandomeffectsasmissingdataandcombiningGibbss

9、amplingandMHalgorithmThehybridalgorithmisusedtoobtainthejointBayesianestimatesofparametersandrandomeffectsStatisticaldiagnosisaboutBayesianmethodvasedeletioninfluencediagnosticsisdevelopedforthejointposteriordistribution