简介：ＡｉｒｂｏｒｎｅＴａｒｇｅｔＳｔａｔｅＥｓｔｉｍａｔｏｒ￥ＨａｏＪｉａｎｋａｎｇ；ＺｈａｎｇＭｉｎｇｌｉａｎ＆ＷｅｎＣｈｕａｎｙｕａｎ（Ｆａｃｕｌｔｙ３０５ｏｆＢｅｉｊｉｎｇＵｎｉｖｅｒｓｉｔｙｏｆＡｅｒｏｎａｕｔｉｃｓ＆ＡｓｔｒｏｎａｕｔｉｃｓＢｅ...

简介：Wecompare13differentaposteriorierrorestimatorsforthePoissonproblemwithlowest-orderfiniteelementdiscretization.Residual-basederrorestimatorscompetewithawiderangeofaveragingestimatorsandestimatorsbasedonlocalproblems.Amongourfivebenchmarkproblemswealsolookontwoexampleswithdiscontinuousisotropicdiffusionandtheirimpactontheperformanceoftheestimators.(SupportedbyDFGResearchCenterMATHEON.)

简介：OneoftheprincipaldisadvantagesofOrthogonalFrequencyDivisionMultiplexing(OFDM)isverysensitivetocarrierfrequencyoffset.Theintegerfrequencyoffsethasnoeffectontheorthogonalityamongthesubcarriers,butitcausesacircularshiftandphaserotationofthereceiveddatasymbolssequence,resultinginaBitErrorRate(BER)of0.5.Inthispaper,anovelintegerfrequencyoffsetestimatorforOFDMisderivedbasedonmaximumlikelihoodestimationtechniqueandexplorationofthedifferentialrelationbetweentwoconsecutiveOFDMdatasymbolsequencesinfrequencydomain.ItsperformanceiscomparedwiththeconventionalmethodbycomputersimulationsfortheadditivewhiteGaussiannoisechannelandamultipathfadingchannel.Simulationresultsshowthattheperformanceoftheproposedestimatorisbetterthantheconventionalestimator.

简介：在它被假定的地方，+σiei(1≤i≤n)被考虑的下列heteroscedastic回归模型Yi=g(xi)那σi[2]=f(ui)，设计指(xi，ui)被知道；非随机，g；f是未知功能。在unobservable骚乱ei形式鞅差别下面，有f的g的小浪评估者的asymptotic规度正在被知道或未知功能被学习。

简介：这份报纸涉及在啤酒杯杯损失下面改进协变性矩阵的评估者的问题。由协变性矩阵的部分Iwasawa坐标，相应风险能被分成三部分。一个人能在加权的二次的损失的加权的矩阵使用信息改进风险的一部分。然而，这份报纸间接地在样品平均数利用信息，复用Iwasawa协调改进风险的其余部分。是值得提及上面的过程能被重复。最后，蒙特卡罗模拟研究被执行验证理论结果。

简介：Inthisarticle,weconsideraclassofkernelquantileestimatorswhichisthelinearcombinationoforderstatistics.Thisclassofkernelquantileestimatorscanberegardedasanextensionofsomeexistingestimators.Theexactmeansquareerrorexpressionforthisclassofestimatorswillbeprovidedwhendataareuniformlydistributed.Theimplementationoftheseestimatorsdependsmostlyonthebandwidthselection.Wethendevelopanadaptivemethodforbandwidthselectionbasedontheintersectionconfidenceintervals(ICI)principle.MonteCarlostudiesdemonstratethatourproposedapproachiscomparativelyremarkable.Weillustrateourmethodwitharealdataset.

简介：AnewmethodalongwithBayesianapproachforestimatingtheparameterinthedistributionfunctionF(x;θ)byusinggroupeddataisdevelopedinthispaper.ThesupportofF(x;θ)isdividedintodisjointedintervalsas-∞=x0

简介：AbstractIntheparametertrackingoftime-varyingsystems,theordinarymethodisweightedleastsquareswiththerectangularwindowortheexponentialwindow.Inthispaperweproposeanewkindofslidingwindowcalledthemultipleexponentialwindow,andthenuseittofittime-varyingGaussianvectorautoregressivemodels.Theasymptoticbiasandcovarianceoftheestimatoroftheparameterfortime-invariantmodelsarealsoderived.Simulationresultsshowthatthemultipleexponentialwindowshavebetterparametertrackingeffectthanrectangularwindowsandexponentialones.

简介：

简介：1.IntroductionTheestimationofpopulationquaillesisofgrestillterestwhenone.isnotpreparedtoassumeaparametricformfortheunderlyingdistribution.Inaddition,quaillesoftenariseasthensturalthingtoestimatewhentheunderlyingdistributionisskewed.LetXIIXZ,’’’)Xubei...

简介：LetX,i.i.d.andY1i.i.d.betwosequencesofrandomvariableswithunknowndistributionfunctionsF(x)andG(y)respectively.X,arecensoredbyY1.InthispaperwestudytheuniformconsistencyoftheKaplan-Meierestimatorunderthecaseey=sup(t:F(t)＜1)＞to=sup(t2G(t)＜1)Thesufficientconditionisdiscussed.

简介：Inthispaper,wediscusstheasymptoticnormalityofthewaveletestimatorofthedensityfunctionbasedoncensoreddata,whenthesurvivalandthecensoringtimesformastationaryα-mixingsequence.Tosimulatethedistributionofestimatorsuchthatitiseasytoperformstatisticalinferenceforthedensityfunction,arandomweightedestimatorofthedensityfunctionisalsoconstructedandinvestigated.Finitesamplebehavioroftheestimatorisinvestigatedviasimulationstoo.更多还原

简介：在这份报纸，我们在LINEX损失功能下面考虑continuousdistribution功能的评价的问题。最好的不变的评估者被获得，是为任何样品尺寸n≥的得最高分的战略的andproved1。

简介：Letfnbeanon-parametrickerneldensityestimatorbasedonakernelfunctionK.andasequenceofindependentandidenticallydistributedrandomvariablestakingvaluesinR.Thegoalofthisarticleistoprovemoderatedeviationsandlargedeviationsforthestatisticsupx∈R|fn(x)-fn(-x)|.

简介：<正>Inthispaper,aclassoffunctionalsofKaplan-Meierestimatorisinvestigated.Countingprocessmartingalemethodsareusedtoshowtheasymptoticnormality,andweestablishameansquareerrorinequalityandaprobabilityinequalityofthemwithouttheassumptionthatF,Garecontinuous,whereF,Garesurvivaltimedistributionandcensoringtimedistributionrespectively.

简介：由使用矩阵的向量方法，我们关于线性限制学习生长曲线模型。在矩阵损失功能和向量损失功能下面，我们在不同类的线性类上为参数的线性评估者的考虑获得必要、足够的条件。

简介：Weinvestigatetheconsistencyandasymptoticnormalityofnearest-neighbordensityestimatorofasampledataprocessbasedonα-mixingassumption.Weextendthecorrespondentresultunderindependentidenticalcases.

简介：在这份报纸，我们在场一个设计方法基于相等的输入骚乱(EID)到的概念为一个线性时间不变的系统拒绝骚乱。一个概括州的观察员(GSO)被用来估计外部骚乱的EID，并且杆任务算法被采用选择GSO的矩阵。模拟和一个旋转速度控制系统的试验性的结果表明我们的方法的有效性。

简介：在一个现代电的驱动程序，转子地面向的控制(RFOC)方法被用来完成好表演和适当短暂回答。在这个方法，转子流动的空间向量来由转子抵抗价值方便。转子抵抗是根据马达速度和房间温度改变变化的重要参数之一。在这份报纸，一个新联机评价方法被利用由使用沃尔什功能领域获得转子抵抗。沃尔什功能是在piecewise常数基础功能(PCBF)的最适用的功能之一解决动态方程。在另一方面，一个不可分的运作的矩阵被用来简化计算算法的过程和速度。模拟结果证明建议方法能够在要用体力地与注射噪音在对比估计转子抵抗的时间间隔上在一台电的机器解决动态方程。

简介：