简介：Inthesystemofm(m≥2)seeminglyunrelatedregressions,weshowthattheGauss-Markovestimator(GME)ofanyregressioncoefficientshasuniquesimplifiedform,whichexactlyequalstotheonestepcovariance-adjustedestimatoroftheregressioncoefficients,andhenceweconcludethatforanyfinitek≥2thek-stepcovariance-adjustedestimatordegeneratestotheone-stepcovariance-adjustedestimatorandthecorrespondingtwo-stageAitkenestimatorhasexactlyonesimplifiedform.Also,theuniquesimplifiedexpressionoftheGMEisjusttheestimatorpresentedintheTheorem1ofWang’work[1988].Anewestimateofregressioncoefficientsinseeminglyunrelatedregressionsystem,ScienceinChina,SeriesA10,1033-1040].

简介：Thispapershowsthemethodofestimatingspatiotemporaldistributionofpedestriansbyusingwatchcameras.Weestimatethedistributionwithouttrackingtechnology,withpedestrian’sprivacyprotectedandinUmedaundergroundmall.Latelyspatiotemporaldistributionofpedestrianshasbeingincreasinglyimportantinthefieldofurbanplanning,disasterpreventionplanning,marketingandsoon.Althoughmanyresearchershavetriedtocapturetheinformationoflocationasdealingwithsomesensors,someproblemsstillremain,suchastheinvestmentofsensors,therestrictionofthenumberofpeoplewhohasthedevicetheyareabletocapture.Fromsuchbackground,wedevelopanoriginallabellingalgorithmandestimatethespatiotemporaldistributionofpedestriansandtheinformationofthepassingtimeandthedirectionofpedestriansfromsequentialimagesofawatchcamera.

简介：中国以社会经济的开发，工业结构，自然资源天赋，和工艺的前进在它的区域之中有巨大的差别。这些差别在中国创造了在区域之间的复杂连接。在这研究，在严肃模型和地点商技术之上造，我们开发一个扇区特定的模型估计内部省的贸易流动，它是为做一张多地区性的输入产量表格的底。在模型，我们与更大的intra扇区输入与更少的intra扇区输入把部门与那些区分开来，并且当时，假设以前的部门趋于在区域之中竞争后者趋于在区域之中合作。然后，我们把内部地区性的贸易评价的这个新方法用于三个部门：食物和烟草，熔炼并且处理的金属，和电的设备。结果证明带宽的那种选择在对内部地区性的贸易的评价以后有重要影响。贸易流动是随带宽的增加散布的更多。作为结果，带宽反映地理活动的空间集中，它应该为不同工业是可区分的。我们断定扇区特定的空间模型能增加内部地区性的贸易流动的估计的可靠。

简介：Background:Remotesensing-basedinventoriesareessentialinestimatingforestcoverintropicalandsubtropicalcountries,wheregroundinventoriescannotbeperformedperiodicallyatalargescaleowingtohighcostsandforestinaccessibility(e.g.REDDprojects)andaremandatoryforconstructinghistoricalrecordsthatcanbeusedasforestcoverbaselines.Giventheconditionsofsuchinventories,thesurveyareaispartitionedintoagridofimagerysegmentsofpre-fixedsizewheretheproportionofforestcovercanbemeasuredwithinsegmentsusingacombinationofunsupervised(automatedorsemi-automated)classificationofsatelliteimageryandmanual(i.e.visualon-screen)enhancements.Becausevisualon-screenoperationsaretimeexpensiveprocedures,manualclassificationcanbeperformedonlyforasampleofimagerysegmentsselectedatafirststage,whileforestcoverwithineachselectedsegmentisestimatedatasecondstagefromasampleofpixelsselectedwithinthesegment.Becauseforestcoverdataarisingfromunsupervisedsatelliteimageryclassificationmaybefreelyavailable(e.g.Landsatimagery)overtheentiresurveyarea(wall-to-walldata)andarelikelytobegoodproxiesofmanuallyclassifiedcoverdata(sampledata),theycanbeadoptedassuitableauxiliaryinformation.Methods:Thequestionishowtochoosethesampleareaswheremanualclassificationiscarriedout.Wehaveinvestigatedtheefficiencyofone-per-stratumstratifiedsamplingforselectingsegmentsandpixels,wheretocarryoutmanualclassificationandtodeterminetheefficiencyofthedifferenceestimatorforexploitingauxiliaryinformationattheestimationlevel.Theperformanceofthisstrategyiscomparedwithsimplerandomsamplingwithoutreplacement.Results:OurresultswereobtainedtheoreticallyfromthreeartificialpopulationsconstructedfromtheLandsatclassification(forest/nonforest)availableatpixellevelforastudyarealocatedincentralItaly,assumingthreelevelsoferrorratesoftheuns

简介：OnthebasisofthepreviousstudiesofthelayeredcrustalmodelintheYutianarea,combinedwiththefieldGPScontinuousobservationdata,weroughlyestimatetheviscouscoefficientofeachlayer.Withtheviscoelastichorizontallayermodel,wecalculatetheviscoelasticco-seismicCoulombstresschangecausedbytheYutianM_S7.3earthquakes2008and2014respectively.BasedontheCoulombstresschange,usingthecalculationmethodof"direct"aftershockfrequency,wecomeupwiththetheoreticalearthquakefrequencydirectlyrelatedtothemainshockandtheco-seismicCoulombstresschangeinthestudyarea.Thenweputforwardamethod,basedonthecomparisonoftheoreticalandactualearthquakefrequencyorthecomparisonbetweentheoreticalandpracticalearthquakefrequency-distancedecaycurvefittingresiduals,toestimatethemagnitudeofamaximumsequentearthquake,directlyrelatedtothemainshockco-seismicCoulombstresschange.Resultscalculatedbydifferentmethodsshowthatthemaximumfollow-upearthquakemagnitudecausedbythecoseismicCoulombstresschangeliesfromM_S7.2toM_S7.5followingYutianM_S7.3earthquakein2008;butthatofthe2014YutianM_S7.3earthquakeisM_S6.3.TheformerisveryclosetotheYutianM_S7.3earthquakein2014.Becauseofthesamemagnitude,relativelyclosespatialdistance,shorttimeinterval,thesameregionoftheexternalforce,thestrongcorrelationbetweentwoseismictectonicandaclearstressinteraction,wethusconsiderthatthetwoYutianM_S7.3earthquakesin2008and2014constituteapairofgeneralizeddoubleshocktypeearthquake.Thisisconsistentwiththesequencetypecharacteristicofpast"doubleshock"earthquakesintheregion.Inthispaper,theinfluenceofthemagnitudelowerlimitandtheb-valueintherelationshipofG-Rontheresultsisdiscussed.Asaresult,whentheviscoelasticcoseismicCoulombstressvariationisdetermined,thelowerlimitofmagnitudehaslittleeffectonthemaximumsequentearthquakemagnitude