Extending Lipschitz functions via random metric partitions

Extending Lipschitz functions via random metric partitions

ID:39459203

大小:307.38 KB

页数:30页

时间:2019-07-03

Extending Lipschitz functions via random metric partitions_第1页
Extending Lipschitz functions via random metric partitions_第2页
Extending Lipschitz functions via random metric partitions_第3页
Extending Lipschitz functions via random metric partitions_第4页
Extending Lipschitz functions via random metric partitions_第5页
资源描述:

《Extending Lipschitz functions via random metric partitions》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、ExtendingLipschitzfunctionsviarandommetricpartitionsJamesR.Lee¤AssafNaorU.C.BerkeleyMicrosoftResearchjrl@cs.berkeley.eduanaor@microsoft.com1IntroductionManyclassicalproblemsingeometryandanalysisinvolvethegluingtogetheroflocalinforma-tiontoproduceacoherentglobalpictur

2、e.Inevitably,thedifficultyofsuchaprocedureliesatthelocalboundary,whereoverlappingviewsofthesamelocalitymustsomehowbemerged.Itisthereforedesirablethattheboundariesbe“smooth,”allowingagracefultransitionfromoneviewpointtothenext.Forinstance,onemaypointtoWhitney’suseofpart

3、itionsofunityinstudyingwhatisnowknownastheWhitneyextensionproblem[36,37].Inthepresentwork,weconsiderwhatisperhapsthemostbasicWhitney-typeextensionproblem,thatofextendingaLipschitzfunctionsothatitremainsLipschitz.Oftensuchamapisextendedbyfirstproducingacoverofthenewdom

4、ain,extendingthemappinglocally,andthengluingtogethertheindividualpieces.Ourmainobservationisthatinmanycases,ifonechoosesarandomcoverfromtherightdistribution,theboundarycanbemade“smooth”onaverage,evenwhenthelocalmapsareindividuallyquitecoarse.Thisinsightleadstotheunifi

5、cation,generalization,andimprovementofmanyknownresults,aswellastonewresultsformanyinterestingspaces.1.1TheLipschitzextensionproblemLet(Y;dY),(Z;dZ)bemetricspaces,andforeveryXµY,denotebye(X;Y;Z)theinfimumoverallconstantsKsuchthateveryLipschitzfunctionf:X!Zcanbeextended

6、toafunctionf˜:Y!Zsatisfyingkf˜kLip·KkfkLip.(IfnosuchKexists,wesete(X;Y;Z)=1).Wealsodefinee(Y;Z)=supfe(X;Y;Z):XµYgandforeveryintegern,en(Y;Z)=supfe(X;Y;Z):XµY;jXj·ng.Estimatinge(Y;Z)isaclassicalandfundamentalproblemthathasattractedalotofattentionduetoitsintrinsicintere

7、standapplicationstogeometryandapproximationtheory.ItisaclassicalfactthatforeverymetricspaceY,e(Y;`1)=1,andKirszbraun’sfamousextensiontheorem[19]statesthatwheneverH1andH2areHilbertspaces,e(H1;H2)=1.Werefertothebooks[3,35]foradetailedaccountofthecasee(Y;Z)=1.Typically,

8、proofsofthefactthate(Y;Z)=1involveshowingthatitispossibletoextendanarbitraryLipschitzfunctiontoanadditionalpointwhilepreservingtheL

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。