简介：Inthispaper,anexponentialinequalityforthemaximalpartialsumsofnegativelysuperadditive-dependent(NSD,inshort)randomvariablesisestablished.Byusingtheexponentialinequality,wepresentsomegeneralresultsonthecompleteconvergenceforarraysofrowwiseNSDrandomvariables,whichimproveorgeneralizethecorrespondingonesofWangetal.[28]andChenetal.[2].Inaddition,somesufficientconditionstoprovethecompleteconvergenceareprovided.Asanapplicationofthecompleteconvergencethatweestablished,wefurtherinvestigatethecompleteconsistencyandconvergencerateoftheestimatorinanonparametricregressionmodelbasedonNSDerrors.

简介：在这份报纸，我们得到H杮甠?灯牥瑡牯吗？

简介：Basedonlefttruncatedandrightcensoreddependentdata,theestimatorsofhigherderivativesofdensityfunctionandhazardratefunctionaregivenbykernelsmoothingmethod.Whenobserveddataexhibitα-mixingdependence,localpropertiesincludingstrongconsistencyandlawofiteratedlogarithmarepresented.Moreover,whenthemodeestimatorisdefinedastherandomvariablethatmaximizesthekerneldensityestimator,theasymptoticnormalityofthemodeestimatorisestablished.

简介：Inthispaper,twofinitedifferencestreamlinediffusion(FDSD)schemesforsolvingtwo-dimensionaltime-dependentconvection-diffusionequationsareconstructed.Stabilityandoptimalordererrorestimati-ionsforconsideredschemesarederivedinthenormstrongerthanL~2-norm.

简介：Consideramultidimensionalrenewalriskmodel,inwhichtheclaimsizes{Xk,k≥1}formasequenceofindependentandidenticallydistributedrandomvectorswithnonnegativecomponentsthatareallowedtobedependentoneachother.Theunivariatemarginaldistributionsofthesevectorshaveconsistentlyvaryingtailsandfinitemeans.Supposethattheclaimsizesandinter-arrivaltimescorrespondinglyformasequenceofindependentandidenticallydistributedrandompairs,witheachpairobeyingadependencestructure.Apreciselargedeviationforthemultidimensionalrenewalriskmodelisobtained.

简介：Undersomemildconditions,wederivetheasymptoticnormalityoftheNadaraya-Watsonandlocallinearestimatorsoftheconditionalhazardfunctionforleft-truncatedanddependentdata.TheestimatorswereproposedbyLiangandOuld-Sai'd[1].TheresultsconfirmtheguessinLiangandOuld-Said[1].

简介：让{X，Xk：k1}是扩大否定地依赖的随机的一个序列有普通分发F令人满意的前的变量>0。让是nonnegative珍视整数的随机的变量，独立于{X，Xk：k1}。在这份报纸，作者为随机的和$S_\tau=\sum\limits_获得必要、足够的条件{n=1}^\tau{X_n}$当随机的数字比被加数有一条更重的尾巴时，有一条一致地变化的尾巴，即，$\frac{{P(X>x)}}{{P(\tau>x)}}\to0$作为x。

简介：

简介：InInternetenvironment,trafficflowtoalinkistypicallymodeledbysuperpositionofON/OFFbasedsources.DuringeachON-periodforaparticularsource,packetsarriveaccordingtoaPoissonprocessandpacketsizes(henceservicetimes)canbegenerallydistributed.Inthispaper,weestablishheavytrafficlimittheoremstoprovidesuitableapproximationsforthesystemunderfirst-infirst-out(FIFO)andwork-conservingservicediscipline,whichstatethat,whenthelengthsofbothON-andOFF-periodsarelightlytailed,thesequencesofthescaledqueuelengthandworkloadprocessesconvergeweaklytoshort-rangedependentreflectingGaussianprocesses,andwhenthelengthsofON-and/orOFF-periodsareheavilytailedwithinfinitevariance,thesequencesconvergeweaklytoeitherreflectingfractionalBrownianmotions(FBMs)orcertaintypeoflongrangedependentreflectingGaussianprocessesdependingonthechoiceofscalingasthenumberofsuperposedsourcestendstoinfinity.Moreover,thesequencesexhibitastatespacecollapse-likepropertywhenthenumberofsourcesislargeenough,whichisakindofextensionofthewell-knownLittle’slawforM/M/1queueingsystem.Theorytojustifytheapproximationsisbasedonappropriateheavytrafficconditionswhichessentiallymeanthattheserviceratecloselyapproachesthearrivalratewhenthenumberofinputsourcestendstoinfinity.

简介：Epidemiologicstudiesuseoutcome-dependentsampling(ODS)schemeswhere,inadditiontoasimplerandomsample,therearealsoanumberofsupplementsamplesthatarecollectedbasedonoutcomevariable.ODSschemeisacost-effectivewaytoimprovestudyefficiency.Wedevelopamaximumsemiparametricempiricallikelihoodestimation(MSELE)fordatafromatwo-stageODSschemeundertheassumptionthatgivencovariate,theoutcomefollowsagenerallinearmodel.Theinformationofbothvalidationsamplesandnonvalidationsamplesareused.Whatismore,weprovetheasymptoticpropertiesoftheproposedMSELE.

简介：Thenumericalsolutionoflargescalemulti-dimensionalconvectiondiffusionequationsoftenrequiresefficientparallelalgorithms.Inthiswork,weconsidertheextensionofarecentlyproposednon-overlappingdomaindecompositionmethodfortwodimensionaltimedependentconvectiondiffusionequationswithvariablecoefficients.Bycombiningpredictor-correctortechnique,modifiedupwinddifferenceswithexplicitimplicitcoupling,themethodunderconsiderationprovidesintrinsicparallelismwhilemaintaininggoodstabilityandaccuracy.Moreover,formulti-dimensionalproblems,themethodcanbereadilyimplementedonamulti-processorsystemanddoesnothavethelimitationonthechoiceofsubdomainsrequiredbysomeothersimilarpredictor-correctororstabilizedschemes.Thesepropertiesofthemethodaredemonstratedinthisworkthroughbothrigorousmathematicalanalysisandnumericalexperiments.

简介：Inthispaper,stochasticglobalexponentialstabilitycriteriafordelayedimpulsiveMarkovianjumpingreaction-diffusionCohen-Grossbergneuralnetworks(CGNNsforshort)areobtainedbyusinganovelLyapunov-Krasovskiifunctionalapproach,linearmatrixinequalities(LMIsforshort)technique,It?formula,Poincar′einequalityandHardy-Poincaréinequality,wheretheCGNNsinvolveuncertainparameters,partiallyunknownMarkoviantransitionrates,andevennonlinearp-Laplacediffusion(p>1).ItisworthmentioningthatellipsoiddomainsinRm(m≥3)canbeconsideredinnumericalsimulationsforthefirsttimeowingtothesyntheticapplicationsofPoincar′einequalityandHardy-Poincar′einequality.Moreover,thesimulationnumericalresultsshowthateventhecorollariesoftheobtainedresultsaremorefeasibleandeffectivethanthemainresultsofsomerecentrelatedliteraturesinviewofsignificantimprovementintheallowableupperboundsofdelays.