Mathematical Statistics - Jun Shaochap7

Mathematical Statistics - Jun Shaochap7

ID:42656883

大小:740.48 KB

页数:72页

时间:2019-09-19

Mathematical Statistics - Jun Shaochap7_第1页
Mathematical Statistics - Jun Shaochap7_第2页
Mathematical Statistics - Jun Shaochap7_第3页
Mathematical Statistics - Jun Shaochap7_第4页
Mathematical Statistics - Jun Shaochap7_第5页
资源描述:

《Mathematical Statistics - Jun Shaochap7》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、Chapter7ConfidenceSetsVariousmethodsofconstructingconfidencesetsareintroducedinthischap-ter,alongwithstudiesofpropertiesofconfidencesets.ThroughoutthischapterX=(X1,...,Xn)denotesasamplefromapopulationP∈P;θ=θ(P)denotesafunctionalfromPtoΘ⊂Rkforafixedintegerk;andC(X)denotesa

2、confidencesetforθ,asetinBΘ(theclassofBorelsetsonΘ)dependingonlyonX.Weadoptthebasicconceptsofconfidencesetsintroducedin§2.4.3.Inparticular,infP∈PP(θ∈C(X))istheconfidencecoefficientofC(X)and,iftheconfidencecoefficientofC(X)is≥1−αforfixedα∈(0,1),thenwesaythatC(X)hassignificancelev

3、el1−αorC(X)isalevel1−αconfidenceset.7.1ConstructionofConfidenceSetsInthissection,weintroducesomebasicmethodsforconstructingconfidencesetsthathaveagivensignificancelevel(orconfidencecoefficient)foranyfixedn.Propertiesandcomparisonsofconfidencesetsaregivenin§7.2.7.1.1Pivotalquan

4、titiesPerhapsthemostpopularmethodofconstructingconfidencesetsistheuseofpivotalquantitiesdefinedasfollows.Definition7.1.AknownBorelfunctionℜof(X,θ)iscalledapivotalquantityifandonlyifthedistributionofℜ(X,θ)doesnotdependonP.NotethatapivotalquantitydependsonPthroughθ=θ(P).Ap

5、ivotalquantityisusuallynotastatistic,althoughitsdistributionisknown.4714727.ConfidenceSetsWithapivotalquantityℜ(X,θ),alevel1−αconfidencesetforanygivenα∈(0,1)canbeobtainedasfollows.First,findtwoconstantsc1andc2suchthatP(c1≤ℜ(X,θ)≤c2)≥1−α.(7.1)Next,defineC(X)={θ∈Θ:c1≤ℜ(X,θ)

6、≤c2}.(7.2)ThenC(X)isalevel1−αconfidenceset,sinceinfPθ∈C(X)=infPc1≤ℜ(X,θ)≤c2P∈PP∈P=Pc1≤ℜ(X,θ)≤c2≥1−α.NotethattheconfidencecoefficientofC(X)maynotbe1−α.Ifℜ(X,θ)hasacontinuousc.d.f.,thenwecanchooseci’ssuchthattheequalityin(7.1)holdsand,therefore,theconfidencesetC(X)hasc

7、onfidencecoefficient1−α.Inagivenproblem,theremaynotexistanypivotalquantity,ortheremaybemanydifferentpivotalquantities.Whentherearemanypivotalquantities,onehastochooseonebasedonsomeprinciplesorcriteria,whicharediscussedin§7.2.Forexample,pivotalquantitiesbasedonsufficientstat

8、isticsarecertainlypreferred.Inmanycaseswealsohavetochooseci’sin(7.1)basedonsomecriteria.Whenℜ(X,θ)andci’sarechosen,weneedtoc

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。