quadratic approximation for nonconvex penalized estimations with a diverging number of parameters

quadratic approximation for nonconvex penalized estimations with a diverging number of parameters

ID:7298987

大小:224.38 KB

页数:8页

时间:2018-02-10

quadratic approximation for nonconvex penalized estimations with a diverging number of parameters_第1页
quadratic approximation for nonconvex penalized estimations with a diverging number of parameters_第2页
quadratic approximation for nonconvex penalized estimations with a diverging number of parameters_第3页
quadratic approximation for nonconvex penalized estimations with a diverging number of parameters_第4页
quadratic approximation for nonconvex penalized estimations with a diverging number of parameters_第5页
资源描述:

《quadratic approximation for nonconvex penalized estimations with a diverging number of parameters》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、StatisticsandProbabilityLetters82(2012)1710–1717ContentslistsavailableatSciVerseScienceDirectStatisticsandProbabilityLettersjournalhomepage:www.elsevier.com/locate/staproQuadraticapproximationfornonconvexpenalizedestimationswithadivergingnumberofparametersSanginLee,Yongd

2、aiKim,SunghoonKwon∗SeoulNationalUniversity,RepublicofKoreaarticleinfoabstractArticlehistory:Weproposeanapproximatedpenalizedestimator(APE)thatcoversvariousstatisticalReceived8August2011modelsandnonconvexpenaltiesincludingthesmoothlyclippedabsolutedeviation(SCAD)Receivedi

3、nrevisedform14May2012penalty(FanandLi,2001)asaspecialcase.TheAPEachievestheoraclepropertywithaAccepted16May2012divergingnumberofparameterswhichextendstheresultsofKwonetal.(2011).SeveralAvailableonline26May2012numericalstudiesconfirmthetheoreticalresults.©2012ElsevierB.V.

4、Allrightsreserved.Keywords:NonconvexpenaltySCADSmoothlyclippedabsolutedeviationOraclepropertyQuadraticapproximation1.IntroductionVariableselectionisoneofthefundamentalissuesforsparsestatisticalmodeling,andpenalizedestimationshavebeenstudiedaseffectivetoolsforthis.Theleas

5、tabsoluteshrinkageandselectionoperator(LASSO)(Tibshirani,1996)andsmoothlyclippedabsolutedeviation(SCAD)penalties(FanandLi,2001)aretworepresentativeexamples.Tibshirani(1996)gaveaniceoverviewontheLASSOasasimultaneousprocedureofvariableselectionandparameterestimation.Howeve

6、r,theLASSOtendstoselectmorevariablesthannecessaryanditbecomesamaindeficiencyoftheLASSO.FanandLi(2001)pointedoutthattheSCADisanimprovedversionoftheLASSOsinceitachievestheoracleproperty:theasymptoticequivalencebetweenthepenalizedestimatorandtheidealnon-penalizedestimatorob

7、tainedwithtruesignalvariablesonly.Foryears,theoraclepropertyhasbeenstudiedintensivelyundervariousstatisticalmodelsandpenalties:theSCAD(FanandLi,2001;FanandPeng,2004),adaptiveLASSO(Zou,2006),bridge(Huangetal.,2008),adaptiveelasticnet(ZouandZhang,2009)andminimaxconcavepena

8、lty(MCP)(Zhang,2010).Inthispaper,weconsiderthequadraticapproximationframeworkstudiedbyKwonetal.(2011).T

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。