the r book generalized additive models

the r book generalized additive models

ID:7295190

大小:341.13 KB

页数:15页

时间:2018-02-10

the r book generalized additive models_第1页
the r book generalized additive models_第2页
the r book generalized additive models_第3页
the r book generalized additive models_第4页
the r book generalized additive models_第5页
资源描述:

《the r book generalized additive models》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、18GeneralizedAdditiveModelsUptothispoint,continuousexplanatoryvariableshavebeenaddedtomodelsaslinearfunctions,linearizedparametrictransformations,orthroughvariouslinkfunctions.Inallcases,anexplicitorimplicitassumptionwasmadeabouttheparametricformofth

2、efunctiontobefittedtothedata(whetherquadratic,logarithmic,exponential,logistic,reciprocalorwhatever).Inmanycases,however,youhaveoneormorecontinuousexplanatoryvariables,butyouhavenoapriorireasontochooseoneparticularparametricformoveranotherfordescribin

3、gtheshapeoftherelationshipbetweentheresponsevariableandtheexplanatoryvariable(s).Generalizedadditivemodels(GAMs)areusefulinsuchcasesbecausetheyallowyoutocapturetheshapeofarelationshipbetweenyandxwithoutprejudgingtheissuebychoosingaparticularparametri

4、cform.Generalizedadditivemodels(implementedinRbythegamfunction)extendtherangeofapplicationofgeneralizedlinearmodels(glm)byallowingnon-parametricsmoothersinadditiontoparametricforms,andthesecanbeassociatedwitharangeoflinkfunctions.Alloftheerrorfamilie

5、sallowedwithglmareavailablewithgam(binomial,poisson,Gamma,etc.).Indeed,gamhasmanyoftheattributesofbothglmandlm,andtheoutputcanbemodifiedusingupdate.Youcanuseallofthefamiliarmethodssuchasprint,plot,summary,anova,predictandfittedafteraGAMhasbeenfittedtod

6、ata.ThegamfunctionusedinthisbookisinthemgcvpackagecontributedbySimonWood:library(mgcv)TherearemanywaysofspecifyingthemodelinaGAM:allofthecontinuousexplanatoryvariablesx,wandzcanenterthemodelasnon-parametricallysmoothedfunctionslikethis:y~s(x)+s(w)+s(

7、z)Alternatively,themodelcancontainamixofparametricallyestimatedparameters(xandz)andsmoothedvariabless(w):y~x+s(w)+zFormulaecaninvolvenested(two-dimensional)termsinwhichthesmoothings()termshavemorethanoneargument,implyinganisotropicsmooth:y~s(x)+s(z)+

8、s(x,z)Alternativelythesmootherscanhaveoverlappingtermssuchasy~s(x,z)+s(z,w)TheRBook,SecondEdition.MichaelJ.Crawley.©2013JohnWiley&Sons,Ltd.Published2013byJohnWiley&Sons,Ltd.GENERALIZEDADDITIVEMODELS667Theuserhasahighdegreeofcontroloverthewaythatinter

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。