简介:本文提出了求解非线性方程组的一种非精确Broyden方法.该方法是文献[8]中精确Broyden方法的推广.在适当的条件下,我们证明了非精确Broyden方法具有全局收敛性和超线性收敛性.数值实验表明,该方法效果较好.
简介:系统综述了自19世纪开始至今常用的统计相关性的方法,例如Pearson和Spearman相关系数,CorGc和CovGc相关性及距离相关性方法。重点介绍了2011年提出的MIC方法以及由此引发的毁誉参半的大量评述,旨在揭示这一热点领域的研究面貌。该领域不仅受到统计学家的关注,而且受到了分析大样本和异质数据的应用研究领域的学者们的追捧,例如基因组生物学家和网络信息研究者。这些研究者期望在众多已有方法的理解和剖析中更恰当地付诸应用,并提出新的应用问题来推动新的分析方法的创造。
简介:Inthispaper,theLp-convergenceofGrünwaldinterpolationGn(f,x)basedonthezerosofJacobipolynomialsJ(α,β)n(x)(-1<α,β<1)isconsidered.Lp-convergence(0<p<2)ofGrünwaldinterpolationGn(f,x)isprovedforp·Max(α,β)<1.Moreover,Lp-convergence(p>0)ofGn(f,x)isobtainedfor-1<α,β≤0.Therefore,theresultsof[1]and[3-5]areimproved.
简介:Inthispaper,weintroduceaconditionweakerthantheLpdifferentiability,whichwecallCpcondition.Weprovethatifafunctionsatisfiesthisconditionatapoint,thenthereexiststhebestlocalapproximationatthatpoint.WealsogiveanecessaryandsufficientconditionforthatafunctionbeLpdifferentiable.Inaddition,westudytheconvexityofthesetofclusterpointsofthenetofbestappoximationsoff,{Pe(f)}ase→0.
简介:Inthepresentpaper,theauthorsintroduceanewsubclassofp-valentanalyticfunctionswithcomplexorderdefinedontheopenunitdiskU=fz:z2Candjzj<1gandobtaincoefficientinequalitiesforthefunctionsintheseclass.Applicationoftheseresultsforthefunctionsdefinedbytheconvolutionarealsoobtained.
简介:LetGbeagraphwithvertexsetV(G)andedgesetE(G)andletgandfbetwointeger-valuedfunctionsdefinedonV(G)suchthat2k-1≤g(x)≤f(x)forallx∈V(G).LetHbeasubgraphofGwithmkedges.Inthispaperitisprovedthatevery(mg+m-1,mf-m+1)-graphGhas(g,f)-factorizationsrandomlyκ-orthogonaltoHandshownthattheresultisbestpossible.
简介:<正>WestudythesolvabilityoftheCauchyproblem(1.1)-(1.2)forthelargestpossibleclassofinitialvalues,forwhich(1.1)-(1.2)hasalocalsolution.Moreover,wealsostudythecriticalcaserelatedtotheinitialvalueu0,for1<p<∞.