部分變量均值隨組別變動(dòng)情況下正態(tài)總體的統(tǒng)計(jì)推斷.pdf

1、2006屆研究生碩士學(xué)位論文y895792學(xué)校代媽：102鼴學(xué)號(hào)：YS∞124030犖柬吁托天擎部分變量均值隨組別變動(dòng)情況下正態(tài)總體的統(tǒng)計(jì)推斷院專研導(dǎo)系統(tǒng)計(jì)系業(yè)概率論與數(shù)理統(tǒng)計(jì)究方島師研究點(diǎn)完成日期數(shù)理統(tǒng)諍王靜龍教授石世磊2006年5月AbstactR印eatedmeasurement8(RM)analy8i8refertothepooungofobserⅦtions0nacr08ssectionofhouseholds，6rmsjet

2、caverseveraltimeperiod8Theyarewidelyu8edinmaIlyfields，for畎咖ple，the80ci出8ciences，financeandsoonF0rRMdatafromnormaldistribution，theob8ervation8甜eoftentransformedbyanorthogonalmatri)(ByobserviⅡgthech盯act凹istic0ftra璐formedda

3、ta，weputforwardolIrque8tionLetp1randomvector8eries蜀，，Kwhicharedistributedaccordingto吼11tiv掃旺iatenormaldistributionwithcommonandunknowncawiancematr政∑Let捌bepartitioneda：8(砑”，磚’，)，sllchthat砑’isam1vector汪l，，mj=1，，口Letx：’，，x囂

4、beindependentandidellticaldistributionwithunknownmeaⅡvectorp(2)and碟0押，一llJ，，碟■唧，Jbeiidwithunknownmeanvec_torp；，j=1，，口Thi8th船isi8devotedtostudyinghowtoestima七e蘆i，，pr’andcov盯iancematrix∑withp：u、啊fyingwhencovariancematrixEi

5、sgeneralSixaspect8ofworkareconsideredThe￡rstaspect醇V鶴thepracticalbackground0fque8tionThe8econdgivesma對(duì)munlikeHhoode8tim砒ionofparameter8bycon—ditionaldistributionandregre88ionmethodThethirddi8cussessomepropositionsofpar咖e

6、terestimationThefourthgi懈1ikelihoodratiote8tstati8ticforte8tingpl叫==pr’The觚hconsiderscovari8皿cetestwhi西i8putforw盯dinch印ter0neThehi出asmpkis百Ven肫刪。礎(chǔ)n們p^瑚esj鼬peatedMe刪Tements，M商叫mLikelih00dEstimat嗵Orthogonaltransform，Me8nch