weighted semiparametric estimation in regression analysis with missing covariate data

weighted semiparametric estimation in regression analysis with missing covariate data

ID:7304954

大小:2.03 MB

页数:14页

时间:2018-02-11

weighted semiparametric estimation in regression analysis with missing covariate data_第1页
weighted semiparametric estimation in regression analysis with missing covariate data_第2页
weighted semiparametric estimation in regression analysis with missing covariate data_第3页
weighted semiparametric estimation in regression analysis with missing covariate data_第4页
weighted semiparametric estimation in regression analysis with missing covariate data_第5页
资源描述:

《weighted semiparametric estimation in regression analysis with missing covariate data》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库

1、WeightedSemiparametricEstimationinRegressionAnalysisWithMissingCovariateDataC.Y.WANG,SuojinWANG,Lue-PingZHAO,andShyh-TyanOuThisarticleinvestigatesestimationoftheregressioncoefficientsinanassumedmeanfunctionwhencovariatesonsomesubjectsaremissing.WeexaminetheperformanceofaHorvitzan

2、dThompson(1952)-typeweightedestimatorbyusingdifferentestimatesoftheselectionprobabilities,whichmaybetreatedasnuisanceparameters(oranuisancefunction).Inparticular,weinvestigatethepropertiesoftheestimateoftheregressionparameterswhentheselectionprobabilitiesareestimatedbykernelsmoot

3、hers.Wepresentlargesampletheoryforthenewestimatorandconductsimulationstudiescomparingtheproposedestimatortothemaximumlikelihoodestimatorandmultipleimputationundervariousmodelassumptionsanddifferentmissingnessmechanisms.Inaddition,weprovidetworealexamplesthatmotivatethisinvestigat

4、ion.KEYWORDS:Case-controlstudy;Estimatingequation;Kernelsmoother;Logisticregression;Measurementerrormodel;Missingatrandom.1.INTRODUCTIONorbylinearregressiononZ'sandY's(Buck1960)fromthecompletecases.Rubin(1978,1987)proposedmultipleTheanalysisofmissingcovariatedataisapracticalimput

5、ationasasolutiontotheproblem.TheEMalgorithmandinterestingproblemintheregressionanalysisofa(Dempster,Laird,andRubin1977)givesausefultoolforunivariateresponse,Y,onavectorofcovariates(X,Z),maximumlikelihoodestimationwhenitrequiresiterativewherethecovariateXmaybemissingpartiallyonasu

6、b-methods.setofrecords.Forexample,inacase-controlstudyofAcloselyrelatedproblemariseswhenXismeasureddietandthyroidcancerconductedinHawaii,39subjectswitherrorandasurrogatevariableforXisavailable.Sev-(totaln=468)hadtheiraveragedailytotaliodinein-eralmethodshavebeenproposed.Forexampl

7、e,Breslowtakesmissing.Anotherexample,illustratedlaterinthisandCain(1988)haveproposedapseudo-conditionallike-article,isacase-controlstudyofesophagusandgastriclihoodmethodfortwo-stagecase-controlstudies.WhencardiacancersinwesternWashingtonstate,whereaboutmissingnessdoesnotdependone

8、itherYorX,Carrollandone-thirdofthesubjec

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。