简介：Weintroduceanewparametrizationofellipticallycontoureddensitiesandstudytheassociatedfamilyofprojected(circular)distributions,Inparticularweinvestigatethetrigonometricmomentsandsomeconvolutionproperties.

简介：Thespectralgradientmethodhasprovedtobeeffectiveforsolvinglarge-scaleunconstrainedoptimizationproblems.Ithasbeenrecentlyextendedandcombinedwiththeprojectedgradientmethodforsolvingoptimizationproblemsonconvexsets.Thiscombinationincludestheuseofnonmonotonelinesearchtechniquestopreservethefastlocalconvergence.Inthisworkwefurtherextendthespectralchoiceofsteplengthtoacceptpreconditioneddirectionswhenagoodpreconditionerisavailable.Wepresentanalgorithmthatcombinesthespectralprojectedgradientmethodwithpreconditioningstrategiestoincreasethelocalspeedofconvergencewhilekeepingtheglobalproperties.Wediscussimplementationdetailsforsolvinglarge-scaleproblems.

简介：Undersuitableconditions,themonotoneconvergenceabouttheprojectediterationmethodforsolvinglinearcomplementarityproblemisprovedandtheinfluenceoftheinvolvedparametermatrixontheconvergencerateofthismethodisinvestigated.

简介：Asubspaceprojectedconjugategradientmethodisproposedforsolvinglargeboundconstrainedquadraticprogramming.Theconjugategradientmethodisusedtoupdatethevariableswithindicesoutsideoftheactiveset,whiletheprojectedgradientmethodisusedtoupdatetheactivevariables.Ateveryiterativelevel,thesearchdirectionconsistsoftwoparts,oneofwhichisasubspacetrumcatedNewtondirection,anotherisamodifiedgradientdirection.Withtheprojectedsearchthealgorithmissuitabletolargeproblems.Theconvergenceofthemethodisprovedandsamenumericaltestswithdimensionsrangingfrom5000to20000aregiven.

简介：Inthispaperweimprovethetwoversionsofthetwo-sidedprojectedquasi-Newtonmethod-onewasproposedbyNocedal&Overtonin[1]andtheotherwasdiscussedinourpreviouspaper,byintroducingthreedifferentmeritfunctionstomakeinexactone-dimensionalsearches.Itisshownthattheseimprovedquasi-Newtonalgorithmshavegainedglobalconvergencepropertywhichisnotpossessedbytheoriginaltwoalgorithms.

简介：Aprojectedsubgradientmethodforsolvingaclassofset-valuedmixedvariationalinequalities(SMVIs)isproposedwhenthemappingisnotnecessarilyLipschitz.Undersomesuitableconditions,itcanbeproventhatthesequencegeneratedbythemethodcanstronglyconvergetotheuniquesolutiontotheproblemintheHilbertspaces.

简介：Weidentifythattheprojecteduncertaintyofthepan-Arcticsea-iceconcentration（SIC）isstronglycoupledwiththeEurasiancirculationintheborealwinter（December-March;DJFM）,basedonasingularvaluedecomposition（SVD）analysisoftheforcedresponseof11CMIP5models.Inthemodelsshowingastrongersea-icedecline,thePolarcellbecomesweakerandthereisananomalousincreaseinthesealevelpressure（SLP）along60°N,includingtheUrals-SiberiaregionandtheIcelandlowregion.ThereisanaccompanyingweakeningofboththemidlatitudewesterlywindsandtheFerrellcell,wheretheSVDsignalsarealsorelatedtoanomalousseasurfacetemperaturewarminginthemidlatitudeNorthAtlantic.IntheMediterraneanregion,theanomalouscirculationresponseshowsadecreasingSLPandincreasingprecipitation.TheanomalousSLPresponsesovertheEuro-AtlanticregionprojectontothenegativeNorthAtlanticOscillation-likepattern.Altogether,pan-ArcticSICdeclinecouldstronglyimpactthewinterEurasianclimate,butweshouldbecautiousaboutthecausalityoftheirlinkage.

简介：七全球联合海洋空气的能力当模特儿复制东方亚洲每月的地面温度和降水气候学在期间19611990被评估。1月和7月气候差别在2050年代和2090年代相对期间19611990在排放情形(SRES)A2和B2情形上在特殊报告下面由七模型的整体投射了然后简短被讨论。和在SRESA2和B2情形下面的相应大气的公司2集中，这些设计随后被用来驱使生物群系模型BIOME3在2050年代和2090年代期间在中国模仿潜在的植被分发。潜在的植被与那些相比在中央、东方的中国极大地向北方在2050年代移动期间系上带子在期间，这被揭示19611990。相反，潜在的植被变化在总体上的西方的中国是细微的。潜在的植被的空间模式在潜在的植被变化对的2050年代，而是范围期间通常在2090年代期间类似于那19611990比2050年代在2090年代期间是更广泛的，特别地在西方的中国。另外，在那里存在潜在的植被的模型依赖者不确定性在2090年代期间在SRESA2情形下面变化，它由于由模型的投射气候变化的scatter。在在SRESA2情形下面的潜在的植被的投射变化35°N向北对35°N的地面温度变化南方并且到地面温度，降水，和大气的公司2集中的联合变化在2090年代期间可归因。

简介：Fortheimprovedtwo-sidedprojectedquasi-Newtonalgorithms,whichwerepresentedinPartI,weproveinthispaperthattheyarelocallyone-steportwo-stepsuperlinearlyconvergent.Numericaltestsarereportedthereafter.ResultsbysolvingasetoftypicalproblemsselectedfromliteraturehavedemonstratedtheextremeimportanceofthesemodificationsinmakingNocedal&Overton’soriginalmethonpractical.Furthermore,theseresultsshowthattheimprovedalgoritnmsareverycompetitiveincomparisonwithsomehighlypraisedsequentialquadraticprogrammingmethods.