简介：Inthispaper,wediscusstherelationbetweenthepartialsumsofJacobiserieronanellipticregionandthecorrespondingpartialsumsofFourierseries.FromthiswederiveapreciseapproximationformulabythepartialsumsofJacobiseriesonanellipticregion.

简介：Inthispaperwepresentanonmonotonetrustregionmethodfornonlinearleastsquaresproblemswithzero-residualandproveitsconvergenceproperties.Theextensivenumericalresultsarereportedwhichshowthatthenonmonotonetrustregionmethodisgenerallysuperiortotheusualtrustregionmethod.

简介：信任区域方法是强大、有效的优化方法。Theconic模型方法比标准二次底的方法是有在每次重复可得到的更多的信息的方法的一种新类型。上述二个方法的优点能被联合为抑制优化形成一个更强大的方法。我们的方法的信任区域潜水艇问题最小化一个圆锥形的功能题目到线性化的限制和信任区域肯定。同时，新算法仍然拥有柔韧的全球性质。在标准条件下面的新算法的全球集中被建立。

简介：Thispaperpresentsatrustregiontwo-phasemodelalgorithmforsolvingtheequalityandboundconstratinednonlinearoptimizationproblem.Aconceptofsubstationarypointisgiven.Undersutableassumptions.theglobalconvergenceofthisalgorithmisprovedwithoutassumingthelinearindependenceofthegradientofactiveconstraints.Anumericalexampleisalsopresented.

简介：Inthispaper,wepresentanefficientapproachforunsupervisedsegmentationofnaturalandtexturalimagesbasedontheextractionofimagefeaturesandafastactivecontoursegmentationmodel.Weaddresstheproblemoftextureswhereneitherthegray-levelinformationnortheboundaryinformationisadequateforobjectextraction.Thisisoftenthecaseofnaturalimagescomposedofbothhomogeneousandtexturedregions.Becausetheseimagescannotbeingeneraldirectlyprocessedbythegray-levelinformation,weproposeanewtexturedescriptorwhichintrinsicallydefinesthegeometryoftexturesusingsemi-localimageinformationandtoolsfromdifferentialgeometry.Then,weusethepopularKullback-Leiblerdistancetodesignanactivecontourmodelwhichdistinguishesthebackgroundandtexturesofinterest.Theexistenceofaminimizingsolutiontotheproposedsegmentationmodelisproven.Finally,atexturesegmentationalgorithmbasedontheSplit-Bregmanmethodisintroducedtoextractmeaningfulobjectsinafastway.Promisingsyntheticandreal-worldresultsforgray-scaleandcolorimagesarepresented.

简介：在这篇论文我们在场为解决LC1的一个filter-trust-region算法使用第二Dini的非强迫的优化问题上面的方向性的衍生物。我们在合理假设下面建立算法的全球集中。

简介：Aclassoftrustregionmethodstorsolvinglinearinequalityconstrainedproblemsispropo6edinthispaper.Itisshownthatthealgorithmisofglobalconvergence.Thealgorithmusesaversionofthetwo-sldedprojectionandthestrategyoftheunconstrainedtrustregionmethods.Itkeepsthegoodconvergencepropertiesoftheunconstrainedcaseandhasthemeritsoftheprojectionmethod.Insomesense,ouralgorithmcanberegardedasanextensionandimprovementoftheprojectedtypealgorithm.

简介：Thispaperproposesanonmonotonicbacktrackingtrustregionalgorithmviabilevellinearprogrammingforsolvingthegeneralmulticommodityminimalcostflowproblems.Usingthedualitytheoryofthelinearprogrammingandconvextheory,thegeneralizeddirectionalderivativeofthegeneralmulticommodityminimalcostflowproblemsisderived.Theglobalconvergenceandsuperlinearconvergencerateoftheproposedalgorithmareestablishedundersomemildconditions.

简介：InthispaperweprovethataclassoftrustregionmethodspresentedinpartⅠissuperlinearlyconvergent.Numericaltestsarereportedthereafter.Resultsbysolvingasetoftypicalproblemsselectedfromliteratureshavedemonstratedthatouralgorithmiseffective.

简介：Inthispaper,wehaveinvestigatedsecondHankeldeterminantsandFeketeSzegoinequalitiesforsomesubclassesofBi-univalentfunctionswithrespecttosymmetricandConjugatepointswhicharesubordinatetoashellshapedregionintheopenunitdisc△.