線性約束下縱向數(shù)據(jù)半?yún)?shù)模型的漸近性質(zhì).pdf

1、山東大學碩士學位論文線性約束下縱向數(shù)據(jù)半?yún)?shù)模型的漸近性質(zhì)姓名：崔文艷申請學位級別：碩士專業(yè)：概率論與數(shù)理統(tǒng)計指導教師：林路20090915ABSTRACTLongitudinaldataisreferredtodatainwhichthesamesampleorindividualismeasuredrepeatedlythroughtimeorspace，SOitcombineselementsofcross—stionaldata

2、andtime—seriesdataTheprominentadvantageoflongitudinaldataisthatitcananalyzeefftivelythechangeofindividualsovertimeHoweveritiscomplicatedbyrepeatedmeasurementwithinindividualsandvariationamongindividualsThemainmethodinstu

3、dyinglongitudinaldataareparametricmethod，nonparametricmethodandsemimetricmethodThemaincontentofthispaperasfollows：1Considerthefollowingsemiparametricregressionmodelforlongitudinaldataunderlinearrestrictions：＼y口=xjpgQq、)e

4、口【R盧=rareobtainedbyusingtheleastsquaresandusualnonparametricweightfunctionmethod，Theasymptoticnormalityoftheestimatorof岱andtheoptimalconvergencerateoftheestimatorofg()areprovedunderthesuitablecondition，andthestrongconver

5、gencerateoftheestimatorofpisgivenunderthesuitablecondition2Considerthefollowingsemiparametricregressionmodelforlongitudinaldata均=吻‘盧g(西)勺Theestimatorsof盧andg()areobtainedbyusingtheleastsquaresandusual(2)nonparametricweig

6、htfunctionmethod，Theasymptoticnormalityoftheestimatorof盧andtheoptimalconvergencerateoftheestimatorofg()areprovedunderthesuitablecondition，andthestrongconvergencerateoftheestimatorof盧andtheuniformstrongconvergencerateofth