generalized cross-validation as a method for choosing a good ridge parameter

generalized cross-validation as a method for choosing a good ridge parameter

ID:7288832

大小:1017.58 KB

页数:10页

时间:2018-02-10

generalized cross-validation as a method for choosing a good ridge parameter_第1页
generalized cross-validation as a method for choosing a good ridge parameter_第2页
generalized cross-validation as a method for choosing a good ridge parameter_第3页
generalized cross-validation as a method for choosing a good ridge parameter_第4页
generalized cross-validation as a method for choosing a good ridge parameter_第5页
资源描述:

《generalized cross-validation as a method for choosing a good ridge parameter》由会员上传分享,免费在线阅读,更多相关内容在工程资料-天天文库

1、AmericanSocietyforQualityGeneralizedCross-ValidationasaMethodforChoosingaGoodRidgeParameterAuthor(s):GeneH.Golub,MichaelHeathandGraceWahbaSource:Technometrics,Vol.21,No.2(May,1979),pp.215-223Publishedby:AmericanStatisticalAssociationandAmericanSocietyforQualityStableURL:http://www.jstor.org/stabl

2、e/1268518.Accessed:16/05/201307:52YouruseoftheJSTORarchiveindicatesyouracceptanceoftheTerms&ConditionsofUse,availableat.http://www.jstor.org/page/info/about/policies/terms.jsp.JSTORisanot-for-profitservicethathelpsscholars,researchers,andstudentsdiscover,use,andbuilduponawiderangeofcontentinatrus

3、teddigitalarchive.Weuseinformationtechnologyandtoolstoincreaseproductivityandfacilitatenewformsofscholarship.FormoreinformationaboutJSTOR,pleasecontactsupport@jstor.org..AmericanStatisticalAssociationandAmericanSocietyforQualityarecollaboratingwithJSTORtodigitize,preserveandextendaccesstoTechnome

4、trics.http://www.jstor.orgThiscontentdownloadedfrom158.182.168.57onThu,16May201307:52:14AMAllusesubjecttoJSTORTermsandConditionsTECHNOMETRICS?,VOL.21,NO.2,MAY1979GeneralizedCross-ValidationasaMethodforChoosingaGoodRidgeParameterGeneH.GolubMichaelHeathDepartmentofComputerScienceComputerSciencesDiv

5、isionStanfordUniversityOakRidgeNationalLaboratoryStanford,CA94303OakRidge,TN37830GraceWahbaDepartmentofStatisticsUniversityofWisconsinMadison,WI53705Considertheridgeestimate:(A)for3inthemodely=Xf+E,e~3(0,a21),a2unknown,0(X)=(XTX+nXI)-XTy.Westudythemethodofgeneralizedcross-validation(GCV)forchoosi

6、ngagoodvalueXforX,fromthedata.TheestimateXistheminimizerofV(A)givenby-V({)-

7、I(-A(X))y2/Trace(A(X)),whereA(X)=X(X7X+nXI)-'X.Thisestimateisarotation-invariantversionofAllen'sPRESS,orordinarycross-validation.Thisestimatebehaveslikeariskimprovementestimator,butdoesnotrequireanestimateofa2,socanbeuse

8、dwhenn-pissmall,orevenifp>nincertaincases.TheGCVmethodcanalsobeusedinsubsetselectionandsingularvaluetruncationmethodsforregression,andeventochoosefromamongmixturesofthesemethods.KEYWORDS=(XTX)-1XTy.(SeeBerger[8],Thiste

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。