1991 Quadratic programming with one negative eigenvalue is NP-hard

1991 Quadratic programming with one negative eigenvalue is NP-hard

ID:39908137

大小:434.27 KB

页数:8页

时间:2019-07-14

1991 Quadratic programming with one negative eigenvalue is NP-hard_第1页
1991 Quadratic programming with one negative eigenvalue is NP-hard_第2页
1991 Quadratic programming with one negative eigenvalue is NP-hard_第3页
1991 Quadratic programming with one negative eigenvalue is NP-hard_第4页
1991 Quadratic programming with one negative eigenvalue is NP-hard_第5页
资源描述:

《1991 Quadratic programming with one negative eigenvalue is NP-hard》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、QuadraticProgrammingwithOneNegativeEigenvalueIsNP-HardPANOSM.PARDALOS’andSTEPHENA.VAVASIS2*‘DepartmentofComputerScience,PennsylvaniaStateUniversity,UniversityPark,PA16802,U.S.A.;‘DepartmentofComputerScience,CornellUniversity,Ithaca,NY14853,U.S.A.(Received:6Februar

2、y1991)Abstract.WeshowthattheproblemofminimizingaconcavequadraticfunctionwithoneconcavedirectionisNP-hard.Thisresultcanbeinterpretedasanattempttounderstandexactlywhatmakesnonconvexquadraticprogrammingproblemshard.Sahniin1974[8]showedthatquadraticprogram-mingwithane

3、gativedefinitequadraticterm(nnegativeeigenvalues)isNP-hard,whereasKozlov,TarasovandHacijan[2]showedin1979thattheellipsoidalgorithmsolvestheconvexquadraticproblem(nonegativeeigenvalues)inpolynomialtime.ThisreportshowsthatevenonenegativeeigenvaluemakestheproblemNP-h

4、ard.Keywords.Globaloptimization,quadraticprogramming,NP-hard.1.IntroductionNonlinearoptimizationincludesmanysubclassesofproblems.Uptodate,thebestmethodsseemnaturallytoexistforconvexprogrammingproblems;see,forexample[2].Nonconvexminimizationproblemsdonotappeartoadm

5、itefficientalgorithms.Theproblemofdesigningalgorithmsthatfindglobalsolutionsisverydifficult,sinceingeneral,therearenolocalcriteriaindecidingwhetheralocaloptimumisglobal[5],[6].Inthispaperweconsiderthecomplexityofaclassofnonconvexquadraticproblems.Thegeneralconcave

6、quadraticproblemhasthefollowingform:minj(x)=ixTQx+cTxs.t.AxGb,wherec,xElRE,AisanwzXnmatrix,bERm,andQisannXnsymmetricnegativesemidefinitematrix.ItiswellknownthatthisproblemisNP-hard[8].Infact,manywellknowncombinatorialoptimizationproblems(e.g.linearO-lprogramming)c

7、anbeformulatedasglobalconcaveminimizationproblems[6].Thesimplestcaseofconcaveprogrammingiswhenthecorrespondingsymmetric*Thisauthor’sworksupportedbytheAppliedMathematicalSciencesProgram(KC-04-02)oftheOfficeofEnergyResearchoftheU.S.DepartmentofEnergyundergrantDE-FGO

8、2-86ER25013.AOOOandinpartbytheNationalScienceFoundation,theAirForceOfficeofScientificResearch,andtheOfficeofNavalResearch,throughNSFgrantDMS8920550.Jour

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。