ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf

ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf

ID:52470625

大小:209.99 KB

页数:14页

时间:2020-03-27

ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf_第1页
ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf_第2页
ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf_第3页
ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf_第4页
ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf_第5页
资源描述:

《ASYMPTOTIC EQUIVALENCE OF EMPIRICALLIKELIHOOD AND BAYESIAN MAP经验地渐近等价性 似然图与贝叶斯图.pdf》由会员上传分享,免费在线阅读,更多相关内容在教育资源-天天文库

1、TheAnnalsofStatistics2009,Vol.37,No.5A,2445–2457DOI:10.1214/08-AOS645cInstituteofMathematicalStatistics,2009ASYMPTOTICEQUIVALENCEOFEMPIRICALLIKELIHOODANDBAYESIANMAPByMarianGrendar´1andGeorgeJudgeBelUniversityandUniversityofCalifornia,BerkeleyInthispaperweareinterestedinempiri

2、callikelihood(EL)asamethodofestimation,andweaddressthefollowingtwoproblems:(1)selectingamongvariousempiricaldiscrepanciesinanELframe-workand(2)demonstratingthatELhasawell-definedprobabilisticinterpretationthatwouldjustifyitsuseinaBayesiancontext.Usingthelargedeviationsapproach

3、,aBayesianlawoflargenumbersisdevelopedthatimpliesthatELandtheBayesianmaximumapos-terioriprobability(MAP)estimatorsareconsistentundermisspeci-ficationandthatELcanbeviewedasanasymptoticformofMAP.Estimatorsbasedonotherempiricaldiscrepanciesare,ingeneral,inconsistentundermisspecifi

4、cation.1.Introduction.Owen’sempiricallikelihood(EL)theorem([30]and[31])providesundertraditionalassumptionsabasisforformingconfidenceregionsformultivariatemeansandparametersinestimatingequations.ThebasicELideaistoproceedasifthesampleX1,X2,...,Xn,drawnfromanunknowndistributionr(

5、x)isi.i.d.andcanbemodeledasamultinomialdistributionbasedontheobservations.InferencefortheJunknownparam-etersisbasedonKestimatingequationsandanonparametriclikelihoodratiostatisticthatasymptoticallyhasachi-squaredistribution.Asaresult,ELisanattractiveorthodoxsemiparametricmetho

6、dofestimationandin-ferencewhosescopehasbeenextendedinseveralproductivedirections(e.g.,see[17]and[41]).BuildingonOwen’sELinsight,thispaperisconcernedwithusingem-arXiv:0908.3397v1[math.ST]24Aug2009piricallikelihoodasamethodofestimation(cf.[3,19,27,31]and[32],amongothers).Throug

7、haBayesianlawoflargenumbers(BLLN,TheoremReceivedFebruary2008;revisedJune2008.1SupportedbyVEGA1/3016/06Grant.AMS2000subjectclassifications.Primary62G05,62C10;secondary60F10.Keywordsandphrases.Maximumnonparametriclikelihood,estimatingequations,Bayesiannonparametricconsistency,Ba

8、yesianlargedeviations,L-divergence,P´olyasam-pling,rightcensoring,Ka

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。