資源描述：
《variable selection in robust joint mean andcovariance model for longitudinal data analysis》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在工程資料-天天文庫。
1、StatisticaSinica24(2014),515-531doi:http://dx.doi.org/10.5705/ss.2011.251VARIABLESELECTIONINROBUSTJOINTMEANANDCOVARIANCEMODELFORLONGITUDINALDATAANALYSISXueyingZheng1,WingKamFung2andZhongyiZhu11FudanUniversityand2TheUniversityofHongKongAbstract:Inlongitudinaldataanalysis,acorrectspeci?cationofthewith
2、in-subjectcovariancematrixcultivatesane?cientestimationformeanregressioncoe?cients.Inthisarticle,weconsiderrobustvariableselectionmethodinajointmeanandcovariancemodel.Weproposeasetofpenalizedrobustgeneralizedestimatingequationstosimultaneouslyestimatethemeanregressioncoe?cients,thegeneral-izedautore
3、gressivecoe?cients,andinnovationvariancesintroducedbythemodi?edCholeskydecomposition.Thesetofestimatingequationsselectimportantcovari-atevariablesinbothmeanandcovariancemodelstogetherwiththeestimatingprocedure.Undersomeregularityconditions,wedeveloptheoraclepropertyoftheproposedrobustvariableselecti
4、onmethod.Finally,asimulationstudyandadetaileddataanalysisarecarriedouttoassessandillustratethesmallsampleper-formance;theyshowthattheproposedmethodperformsfavorablybycombiningtherobustifyingandpenalizedestimatingtechniquestogetherinthejointmeanandcovariancemodel.Keywordsandphrases:Covariancematrix,p
5、enalizedgeneralizedestimatingequa-tion,longitudinaldata,modi?edcholeskydecomposition,robustness,variablese-lection.1.IntroductionLongitudinaldataarisemoreandmorefrequentlyinavarietyofscienti?cdomainsthatseekinsightfulandcomprehensiveresearchinabranchofstatisti-calmodeling.Di?erentfromothertypesofdat
6、a,weoftenassumeindependenceamongdistinctsubjectsbutdependencewithineachsubject;within-subjectcor-relationraisesafundamentalchallengefortheanalysisoflongitudinaldata.LiangandZeger(1986),amilestoneinthedevelopmentofmethodologyforlongitudinaldataanalysis,proposedgeneralizedestimatingequations(GEE)fores
7、timationofgeneralizedlinearregressioncoe?cients.Themainadvantageoftheirmethodisthatevenwhenthewithin-subjectcorrelationistreatedasanuisanceparameterwithanassumedparsimoniousstructure,GEEstillbringsabo
