optimization under unknown constraints

optimization under unknown constraints

ID:7290894

大小:687.69 KB

页数:31页

时间:2018-02-10

optimization under unknown constraints_第1页
optimization under unknown constraints_第2页
optimization under unknown constraints_第3页
optimization under unknown constraints_第4页
optimization under unknown constraints_第5页
资源描述:

《optimization under unknown constraints》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、OptimizationUnderUnknownConstraints*UniversityPressScholarshipOnlineOxfordScholarshipOnlineBayesianStatistics9JoséM.Bernardo,M.J.Bayarri,JamesO.Berger,A.P.Dawid,DavidHeckerman,AdrianF.M.Smith,andMikeWestPrintpublicationdate:2011PrintISBN-13:9780199694587PublishedtoOxfordScholarshipOnline:Ja

2、nuary2012DOI:10.1093/acprof:oso/9780199694587.001.0001OptimizationUnderUnknownConstraints*RobertB.GramacyHerbertK.H.LeeDOI:10.1093/acprof:oso/9780199694587.003.0008AbstractandKeywordsOptimizationofcomplexfunctions,suchastheoutputofcomputersimulators,isadifficulttaskthathasreceivedmuchattent

3、ionintheliterature.Alessstudiedproblemisthatofoptimizationunderunknownconstraints,i.e.,whenthesimulatormustbeinvokedbothtodeterminethetypicalreal‐valuedresponseandtodetermineifaconstrainthasbeenviolated,eitherforphysicalorpolicyreasons.WedevelopastatisticalapproachbasedonGaussianprocessesan

4、dBayesianlearningtobothapproximatetheunknownfunctionandestimatetheprobabilityofmeetingtheconstraints.Anewintegratedimprovementcriterionisproposedtorecognizethatresponsesfrominputsthatviolatetheconstraintmaystillbeinformativeaboutthefunction,andthuscouldpotentiallybeusefulintheoptimization.T

5、henewcriterionisillustratedonsyntheticdata,andonamotivatingoptimizationproblemfromhealthcarepolicy.Page1of31OptimizationUnderUnknownConstraints*Keywords:ConstrainedOptimization,SurrogateModel,GaussianProcess,SequentialDesign,ExpectedImprovementSummaryOptimizationofcomplexfunctions,suchasthe

6、outputofcomputersimulators,isadifficulttaskthathasreceivedmuchattentionintheliterature.Alessstudiedproblemisthatofoptimizationunderunknownconstraints,i.e.,whenthesimulatormustbeinvokedbothtodeterminethetypicalreal‐valuedresponseandtodetermineifaconstrainthasbeenviolated,eitherforphysicalorp

7、olicyreasons.WedevelopastatisticalapproachbasedonGaussianprocessesandBayesianlearningtobothapproximatetheunknownfunctionandestimatetheprobabilityofmeetingtheconstraints.Anewintegratedimprovementcriterionisproposedtorecognizethatresponsesfrominputsthatvio

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。