资源描述:
《identifying key variables and interactions in statistical models of building energy consumption using regularization》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、Energy83(2015)144e155ContentslistsavailableatScienceDirectEnergyjournalhomepage:www.elsevier.com/locate/energyIdentifyingkeyvariablesandinteractionsinstatisticalmodelsofbuildingenergyconsumptionusingregularization*DavidHsu210S34thStreet,Philadelphia,PA19104,USAartic
2、leinfoabstractArticlehistory:Statisticalmodelscanonlybeasgoodasthedataputintothem.DataaboutenergyconsumptionReceived2September2014continuestogrow,particularlyitsnon-technicalaspects,butthesevariablesareofteninterpretedReceivedinrevisedformdifferentlyamongdisciplines
3、,datasets,andcontexts.Selectingkeyvariablesandinteractionsistherefore9January2015animportantstepinachievingmoreaccuratepredictions,betterinterpretation,andidentificationofkeyAccepted4February2015subgroupsforfurtheranalysis.Availableonline6March2015Thispaperthereforem
4、akestwomaincontributionstothemodelingandanalysisofenergycon-sumptionofbuildings.First,itintroducesregularization,alsoknownaspenalizedregression,forprin-Keywords:cipledselectionofvariablesandinteractions.Second,thisapproachisdemonstratedbyapplicationtoaEnergyconsumpt
5、ionBuildingscomprehensivedatasetofenergyconsumptionforcommercialofficeandmultifamilybuildingsinNewVariableselectionYorkCity.Usingcross-validation,thispaperfindsthatanewly-developedmethod,hierarchicalgroup-Statisticalmodelslassoregularization,significantlyoutperformsrid
6、ge,lasso,elasticnetandordinaryleastsquaresap-proachesintermsofpredictionaccuracy;developsaparsimoniousmodelforlargeNewYorkCitybuildings;andidentifiesseveralinteractionsbetweentechnicalandnon-technicalparametersforfurtheranalysis,policydevelopmentandtargeting.Thismeth
7、odisgeneralizabletootherlocalcontexts,andislikelytobeusefulforthemodelingofothersectorsofenergyconsumptionaswell.©2015TheAuthor.PublishedbyElsevierLtd.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).1.Introductionincludingac
8、curatepredictions,theinterpretationofeffects,andidentificationofkeysubgroupsforfurtherphysics-basedmodelingStatisticalmodelsenableempirical