欢迎来到天天文库
浏览记录
ID:32074756
大小:5.83 MB
页数:50页
时间:2019-01-31
《投资组合优化模型的文化算法研究》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、AbstractPortfolioisanimportantpartofmodemfinancewhichprovidesawaytoallocatethecapitaltoseveralriskassetsinordertominimizetheriskormaximizetheprofit.Atpresent,optimizationmethodsandadvancedcomputationaltechnologyare’usedtosolvetheportfoliooptimizationmodels,
2、especiallytheapplicationofintelligentalgorithms.CulturalalgorithmisarecentlyemergingbionicmteIllgentalgorithmthatprovidesauniversalframeworkofthedouble-layerevolution,whileovercomangthedeftcienciesofthetraditionalintelligentalgorithmfortherecessiveexpressio
3、nandacce88ottneexperienceandknowledge.Withthehelpoftheuniquecomputationalframeworkofculturalalgorithmandthegoodparallel,thisthesiseffectivelyimprovesthealgorithmbyusingtheideaofintegrationandappliestheimprovedhybridalgorithmforsolvingthefunctionoptimization
4、andportfolio0pt蛐1zatlonmodels.Themainworkofthisthesisisasfollows:1.ThisthesismakestheparticleswarmoptimizationembedintoculturalalgorithmasitspopulatlonsDaceandbeliefspace,andproposesanewdynamicparticleswarmoptlmlzatmnbasedonculnlraIalgorithm.Asetofdynamicco
5、mmunicationoperationisdesignedtoachievetheinteractionbetweenthetwosDacesinordertoimproveandovercomethedefectthattheparticleswarmoptimizatlon1seasllYtrappedintothelocaloptimum.ExperimentalresultsofseveralBenchmarkfunctionsshowthatthisalgorithmnotonlyhasahigh
6、precisionbutalsohasastrongrobustness,anditisaneffectiveglobaloptimizationalgorithm.Portfoliodecisionfacingagreatcomplicatedcombinatorialoptimizationproblem,andisdealofdataoftherealsecuritymarketisaNP.hardproblem,whichisdifficulttobesolvedbytraditionalalgori
7、thm.Thisthesisintroducesanewdynamicparticleswarmopti叫zatlonbased0nculturalalgorithmforsolvingthemean·VaRmodel,andusesthepenaltyfunctionapproachtotheinequalityconstraintsinthemodel.AnempiricalanalysisresultsshowthattheproblemofportfoliooDtimizationcanbesolve
8、dmorereasonablyandefficientlybythisalgorithm.2.Basedextemalelitearchiveandcrowdingentropystrategy,thisthesisgwesanewmult卜objectiveDarticleswarrnoptimizationbasedonculturalalgorithm.Firstly,itistestedon
此文档下载收益归作者所有