简介：这篇论文的目的是提供错误分析为多范畴支持向量机器(MSVM)classificaton问题。我们为MSVM建立一致集聚途径并且估计错误分类错误。我们这里克服的主要困难是围住偏移量向量。作为结果，当核多项式的度足够大时，我们证实有多项式核的MSVM分类算法总是是有效的。最后，集中和例子的率被给表明主要结果。

简介：影响司烧混合的层温度变化上的飘移被考虑Stokes飘移在混合的层温度并且用的方程导致的一个移流的热运输学期估计海洋并且从一个全球海洋发行量模型(HYCOM)和一个波浪的波浪参数当模特儿(波浪手表III)。维的分析和量的评价方法被进行估计Stokes飘移导致的效果的重要性并且分析它的空间分发和季节的变化特征。到在mid-to-high纬度的混合的层温度变化的Stokes飘移的贡献与吝啬的水流的是可比较的结果表演，和混合的层温度变化的实质的部分被考虑Stokes飘移效果导致。尽管Stokes飘移导致的移流热运输不是为混合的层温度方程的领先的术语，它不能被忽视并且甚至为upperocean温度的模拟在一些区域变得批评。

简介：Thispaperpresentsaheattransfermodelforthehyperthermiatreatmentofcervixcancerusingaintracavitarymicrowaveapplicatorandbasedonwhichthe3-Dfiniteelementsimulationofthetemperaturefieldshavedone,Beforethenthespecificabsorptionrate(SAR)distributionforthetransvaginalprobehavebeenmeasuredinaphantom.Thevariationsoftheparametershavebeeninvestigated.Too,foroptimization.Atlast,thersultsofsimulationarecomparedtothatmeasuredinthephantomandsomeinstructiveconclusionsarepresentedforcriticalapplication.

简介：LaizhouBayanditsadjacentwatersareofgreatimportancetoChina’smarineoilandgasdevelopment.Itisthereforecrucialtoestimatereturn-periodvaluesofmarineenvironmentalvariablesinthisregiontoensurethesafetyandsuccessofmaritimeengineeringandmaritimeexploration.Inthisstudy,weusednumericalsimulationstoestimateextremewaveheight,seacurrentvelocityandsea-levelheightinwesternLaizhouBay.Theresultsshowthatthesea-levelrisestartsatthemouthofthebay,increasestowardwest/southwest,andreachesitsmaximuminthedeepestbasinofthebay.The100-yearreturn-periodvaluesofsealevelrisecanreach3.4–4.0minthewesternbay.TheelevationofthewesternpartoftheQingdongOilFieldwouldremainabovetheseasurfaceduringextremelowsealevel,whiletherestoftheoilfieldwouldbe1.6–2.4mbelowtheseasurface.Thereturn-periodvalueofwaveheightisstronglyaffectedbywaterdepth;infact,itsspatialdistributionissimilartotheisobath’s.The100-yearreturn-periodvaluesofeffectivewaveheightcanbe6morhigherinthecentralbayandbemorethan1mintheshallowwaternearshore.The100-yearreturn-periodvaluesofcurrentvelocityisabout1.2–1.8ms-1intheQingdongOilField.Theseresultsprovidescientificbasisforensuringconstructionsafetyandreducingconstructioncost.

简介：Accuratetimedelayestimationisveryimportantinmultichannelnoisesuppressionsystems.Thispaperhasproposedanewmethodtoreducetheeffectofnoiseontheestimationoftimedelaybetweendifferentchannels.Ournewtimedelayalgorithmwillmoreconsiderablyreducetheestimationerror,comparedwiththeconventionalautocorrelationfunctionmethod.

简介：这份报纸为感动一般水准(PARMA)模特儿的原因、可颠倒的周期的autoregressive调查加权的最不绝对的偏差评估者(WLADE)。评估者的Asymptotic规度在一个部分时刻条件下面被导出。模拟研究被给估计建议WLADE的表演。

简介：Anewadaptiveestimatorfordirectsequencespreadspectrum(DSSS)signalsusingfourth-ordercumulantbasedadaptivemethodisconsidered.Thegeneralhigher-orderstatisticsmaynotbeeasilyappliedinsignalprocessingwithtoocomplexcomputation.Basedonthefourth-ordercumulantwith1-Dslicesandadaptivefilters,anefficientalgorithmisproposedtosolvetheproblemandisextendedfornonstationarystochasticprocesses.Inordertoachievetheaccurateparameterestimationofdirectsequencespreadspectrum(DSSS)signals,thefirststepusesthemodifiedfourth-ordercumulanttoreducethecomputingcomplexity.Whilethesecondstepemploysanadaptiverecursivesystemtoestimatethepowerspectruminthefrequencydomain.Inthecaseofinterceptedsignalswithoutlargeenoughdatasamples,theestimatorprovidesgoodperformanceinparameterestimationandwhiteGaussiannoisesuppression.Computersimulationsareincludedtocorroboratethetheoreticaldevelopmentwithdifferentsignal-to-noiseratioconditionsandrecursivecoefficients.

简介：Regularizationmethodisaneffectivemethodforsolvingill-posedequation.Inthispapertheunbiasedestimationformulaofunitweightstandarddeviationintheregularizationsolutionisderivedandtheformulaisverifiedwithnumericalcaseof1000sampledatabyuseofthetypicalill-posedequation,i.e.theFredholmintegrationequationofthefirstkind.

简介：Bytakingtheurbangreenlandastestingobjective,theGeoEye-1fusionimageasbackground,andusingMoran’sindex,thecorrelativityofgreenlandunitshasbeenmeasured.Then,thepresamplingwasconductedbasedonthemeasuredcorrelativity,thesystematicsamplingdesignhasbeencarriedoutattheprecision95%and85%.ByusingofKriginginterpolationmethod,theweightofdifferenttypeofgreenlandhasbeendividedintothestratifications.Throughthestratifiedsampling,theestimationofgreenlandcoveragehasbeenobtained.Theresultsshowthattheurbangreenlandcoverageareais14.81km~2,accountingfor27.73%ofallurbanlandareas.Comparedwiththecommonsamplingmethod,theprecisionofthatincreasedabout3%.Thismethodcanalsobeutilizedinmonitoringdynamicvariationofurbangreenlandwiththerealtimeimages.

简介：

简介：Thedevelopmentofhigh-resolutionremotesensingimagingtechnologyprovidesanewwaytothelarge-scaleestimationofforestcanopydensity.Thetraditionalinversionmethodsforcanopydensityonlyusespectralortopographicalfeaturesofremotesensingimages.However,duetotheexistenceofthedifferentthingwithsamespectrumandthesamethingwithdifferentspectrumphenomena,itisdifficulttoimprovetheestimationaccuracyofcanopydensity.Basedonspectrumandothertraditionalfeatures,thispapercombinestexturefeaturesofremotesensingimagestoestimatecanopydensity.Firstly,thegraylevelco-occurrencematrix(GLCM)texturefeaturesarecomputedusingobjectbasedmethod.Then,theprincipalcomponentanalysis(PCA)methodisappliedincorrelationanalysisanddimensionreductionoftexturefeatures.Finally,spectrumandtopographicalfeaturestogetherwithtexturefeaturesareintroducedintostepwiseregressionmodeltoestimatecanopydensity.Theexperimentalresultsshowedthatcomparedwiththetraditionalmethodonlybasedonspectrumortopographicalfeatures,themethodcombinedwithtexturefeaturesgreatlyimprovedtheestimationaccuracy.Thecoefficientofdetermination(adjustedR~2)increasedfrom0.737to0.805.Theestimationaccuracyincreasedfrom81.03%to84.32%.

简介：Theobservationandestimationofdirectionalspectraofseawavesisoneoftheessentialsubjectsofstudyofoceanicdynamics.Onthebasisoftheirregularlinearwavetheory,estimationmethodsfori/UV,PUVandVVdirectionalwavespectraarederived.ByusingijandPUVdatameasuredin-situ,directionalwavespectraareestimated,meanwhilethevirtuesanddefectsofvariousspectraarecomparied.Thismethodprovideabasisfortheobservationsofseawaves.

简介：在这篇论文，处于多重变化点评价的问题的最少的方形的评估者被学习。这里，在吝啬的移动的ALNQD顺序的动人一般水准的过程被讨论。当变化点的数字被知道时，集中ofchange指评价的率被导出。结果也是真的为，混合的ρ-，混合的φ-，混合的α-，在合适的条件下面联系了并且否定地联系了序列。

简介：InordertoprovideascientificbasisforriceyieldestimationandimprovetheacouracyofyieldestimationinZhejiangProvince,Regionalizationindicesforriceyieldestimationbyremoesensing(RS)intheprovinceweredeterminedbyconsjderingthespecialfeaturesofyieldestimationbyRS,andbasedonanalysisofthenaturalconditionsofZhejiangProvince,Theindicesdeterminedincludedricecroppingsystem,agroclimate,landform,surfacefeaturesturctureandriceyieldlevel,wherericeplantingsystemwasconsideredasthemianone,ThenreionalizationfroriceyieldestimationbyRSwascompletedbyspatialneighboringanalysiswiththeGeographicalInfromationSystem(GIS)technologycombinedwithusigoftreealgorithmTheprovincewasdividedintotworegions,i.e.,thesingle-croppingriceregionwhichwassubdividedinto3regionsincludingthoseinmountainsofnorthewstZhejiang,waternetworkareaofnorthZhejiangandmountainsofsothZhejiang,anddouble-croppingriceregionwhichwassudividedinto5regionsincludingthoseonplainofnorthZhejiang,coastalplainsandhillsofsoutheastZhejiang,Jin-QuBasinofmiddleZhejiang,hillsofeastZhejiang,andhillsandmountainsofnorthewatZhejiang,thisregionaliztiontookthecountybordersastheregionboundaries,kepttheregionsconnectiveandmadetheadministrativregionsintegrityand,then,couldmeettherequirementsofriceyieldestimationbyRS,showingthattheresultswerequitesatisfying.

简介：

简介：Thechlorophyll-aconcentrationisgenerallyoverestimatedforthesouthemChinacoastalwatersifthedefaultalgorithmoftheSeaDASisemployed.Analgorithmisdevelopedforretrievalofchlorophyll-aconcentrationintheZhujiangEstuary,GuangdongProvince,China,byusingsimulatedreflectancedata.ThesimulatedreflectanceiscalculatedcorrespondingtotheSeaWiFSwavelengthbands,viaageneralmodelbyinputtingmeasuredwatercomponents,I.e.,thesuspendedsediment,chlorophyll-a,andyellowsubstance(DOC)concentrationdataof130samples.Empiricalrelationshipsofthechlorophyll-aconcentrationto240differentbandcombinationsareinvestigatedbasedonthesimulatedreflectancedata,andthebandcombination,R5R6/R3R4,isfoundtobetheoptimumoneforthedevelopmentofanalgorithmvalidfortheZhujiangEstuary.Thisalgorithmisthenemployedtodeterminethechlorophyll-aconcentrationfromSeaWiFSdata.TheestimatedconcentrationshaveabetteraccuracythanthoseobtainedfromtheSeaDASdefaultalgorithmwhencomnparedwithseatruthdata.Thenewalgorithmisdemonstratedtoworkwellandisusedtoderiveaseriesofimagemapsofthechlorophyll-aconcentrationdistributionfortheZhujiangEstuaryandadjacentcoastalareas.

简介：在收音机频率鉴定(RFID)系统，backscattered信号对干扰小、容易。RFID标签identi的表演?在干扰情形的阳离子在无差错的情形与那相比被降级。在这份报纸，一个新奇Mahalanobis距离估计(MDE)方法被建议联合估计标签和包错误率的数字(每)。当干扰正在损害观察信息时，MDE方法是由于它完成一个稳定的评价的能力有弹性的错误。建议方法由在观察结果之中考虑所有信息和关联在存在方法上完成显著地提高的精确性。MDE方法基于观察信息的有效decorrelation和分类改进估计表演。而且，PER估计的表演在理论并且通过模拟被分析。它能从分析被结束估计了每不偏、围住变化。模拟证明建议估计以精确性和稳定性超过以前的建议，它在干扰情形为申请使它合适。

简介：Thetraditionalsuper-resolutiondirectionfindingmethodsbasedonsparserecoveryneedtodividetheestimationspaceintoseveraldiscreteanglegrids,whichwillbringthefinalresultsomeerror.Tothisproblem,anovelmethodforwidebandsignalsbysparserecoveryinthefrequencydomainisproposed.Theoptimizationfunctionsarefoundandsolvedbythereceiveddataateveryfrequency,onthisbasis,thesparsesupportsetisobtained,thenthedirectionofarrival(DOA)isacquiredbyintegratingtheinformationofallfrequencybins,andtheinitialsignalcanalsoberecovered.Thismethodavoidstheerrorcausedbysparserecoverymethodsbasedongriddivision,andthedegreeoffreedomisalsoexpandedbyarraytransformation,especiallyithasapreferableperformanceunderthecircumstancesofasmallnumberofsnapshotsandalowsignaltonoiseratio(SNR).

简介：EstimationofThomsen＇sanisotropicparametersisveryimportantforaccuratetime-to-depthconversionanddepthmigrationdataprocessing.Comparedwithothermethods,itismucheasierandmorereliabletoestimateanisotropicparametersthatarerequiredforsurfaceseismicdepthimagingfromverticalseismicprofile（VSP）data,becausethefirstarrivalsofVSPdatacanbepickedwithmuchhigheraccuracy.Inthisstudy,wedevelopedamethodforestimatingThomsen＇sP-waveanisotropicparametersinVTImediausingthefirstarrivalsfromwalkawayVSPdata.Modelfirst-arrivaltraveltimesarecalculatedonthebasisofthenear-offsetnormalmoveoutcorrectionvelocityinVTImediaandraytracingusingThomsen＇sP-wavevelocityapproximation.Then,theanisotropicparameters0andearedeterminedbyminimizingthedifferencebetweenthecalculatedandobservedtraveltimesforthenearandfaroffsets.Numericalforwardmodeling,usingtheproposedmethodindicatesthaterrorsbetweentheestimatedandmeasuredanisotropicparametersaresmall.Usingfielddatafromaneight-azimuthwalkawayVSPinTarimBasin,weestimatedtheparameters0andeandbuiltananisotropicdepth-velocitymodelforprestackdepthmigrationprocessingofsurface3Dseismicdata.Theresultsshowimprovementinimagingthecarbonatereservoirsandminimizingthedeptherrorsofthegeologicaltargets.

简介：Thispaperdealswithestimationandtestproceduresforrestrictedlinearerrors-invariables(EV)modelswithnonignorablemissingcovariates.Wedeveloparestrictedweightedcorrectedleastsquares(WCLS)estimatorbasedonthepropensityscore,whichisfittedbyanexponentiallytiltedlikelihoodmethod.Thelimitingdistributionsoftheproposedestimatorsarediscussedwhentiltedparameterisknownorunknown.Totestthevalidityoftheconstraints,weconstructtwotestproceduresbasedoncorrectedresidualsumofsquaresandempiricallikelihoodmethodandderivetheirasymptoticproperties.Numericalstudiesareconductedtoexaminethefinitesampleperformanceofourproposedmethods.