線性模型參數(shù)M估計(jì)的相合性.pdf

1、中國科學(xué)技術(shù)大學(xué)碩士學(xué)位論文線性模型參數(shù)M估計(jì)的相合性姓名：陳建東申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：概率論與數(shù)理統(tǒng)計(jì)指導(dǎo)教師：吳耀華20051101中國科學(xué)技術(shù)大學(xué)碩士學(xué)位論文英文摘要AbstractResearchandpracticehaveshownthat，atypicalwayinlinearregressionmodelsleastsquaresestimate(LSE)，mightbeverybadinsomeconditionsF

2、orrecenthalfcenturystatisticshavefoundmanymethodstoresolveitIn1964，HuberfirstfoundMestimatesofalocationparameterIn1973，healsesearchedM—estimatesinthelinearmodelSincethen，Mestimateshavebeenrespectedandmanyimportantpapersh

3、avepressedInthispaper’weonlyconsiderthelinearregressionmodelandindependentrandomerrors，thatis，yi2x’f口oe』，i=1，2，，Ⅳ，hereXl，X2，，xnareknownrandomvectors，el，e2，，enareerrorterms，130isanunknownregressionparametervectorThiscasei

4、sthemostusefulinstatisticsFortheabovelinearmodel，wesupposePisaconvexfunction，whichwillbeusefulforUStostudyMestimation，WeaimtoexploreforsomesuitablesufficientconditionsofconsistencyandconvergenceofMestimationofregressionp

5、arametersIfthisproblemisworkedout，itisamarkpoleforcomplexmodels，Thus，itisimportanceforthestudyandapplyofMestimationChenXRandZhanLC(Eestimatorinlinearmodel，1996)haveentirelysummarizedrelatedworks，wheremanyprofoundproblems

6、areposed，andconsistencyandconvergenceproblemsaremostbasicandimponantItaffordusmanymaterial，ideasandskillsSOthatwecarlstudythemTherearethreechaptersinthisthesisInchapter1，wesimplyintroducesomeimportantliteratureOnMestimat

7、ioninthelinearregressionmodelInchapter2，weintroduceourmainresultsmoredetailed，herethestrongconsistencyandconvergenceofM—estimationregressionparametersinlinearmodelsareestablished，andsomeresultsarealsoimprovedInchapter3，w