资源描述:
《Very Deep Convolutional Networks》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、VeryDeepConvolutionalNetworksforTextClassificationAlexisConneauHolgerSchwenkYannLeCunFacebookAIResearchFacebookAIResearchFacebookAIResearchaconneau@fb.comschwenk@fb.comyann@fb.comLo¨ıcBarraultLIUM,UniversityofLeMans,Franceloic.barrault@univ-lemans.frAbstractterestintheresearch
2、communityandtheyaresys-tematicallyappliedtoallNLPtasks.However,ThedominantapproachformanyNLPwhiletheuseof(deep)neuralnetworksinNLPtasksarerecurrentneuralnetworks,inpar-hasshownverygoodresultsformanytasks,itticularLSTMs,andconvolutionalneuralseemsthattheyhavenotyetreachedthele
3、veltonetworks.However,thesearchitecturesoutperformthestate-of-the-artbyalargemargin,arerathershallowincomparisontotheasitwasobservedincomputervisionandspeechdeepconvolutionalnetworkswhichhaverecognition.pushedthestate-of-the-artincomputervi-Convolutionalneuralnetworks,inshort
4、Con-sion.Wepresentanewarchitecture(VD-vNets,areverysuccessfulincomputervision.InCNN)fortextprocessingwhichoperatesearlyapproachestocomputervision,handcrafteddirectlyatthecharacterlevelandusesfeatureswereused,forinstance“scale-invariantonlysmallconvolutionsandpoolingoper-featu
5、retransform(SIFT)”(Lowe,2004),followedations.Weareabletoshowthattheper-bysomeclassifier.ThefundamentalideaofCon-formanceofthismodelincreaseswiththevNets(LeCunetal.,1998)istoconsiderfeaturedepth:usingupto29convolutionallayers,extractionandclassificationasonejointlytrainedwerepor
6、timprovementsoverthestate-of-task.Thisideahasbeenimprovedovertheyears,the-artonseveralpublictextclassificationinparticularbyusingmanylayersofconvolutionstasks.Tothebestofourknowledge,thisisandpoolingtosequentiallyextractahierarchicalthefirsttimethatverydeepconvolutionalrepresen
7、tation(ZeilerandFergus,2014)ofthein-netshavebeenappliedtotextprocessing.put.Thebestnetworksareusingmorethan150layersasin(Heetal.,2016a;Heetal.,2016b).1IntroductionarXiv:1606.01781v2[cs.CL]27Jan2017ManyNLPapproachesconsiderwordsasba-Thegoalofnaturallanguageprocessing(NLP)issic
8、units.Animportantstepwastheintroductiontoprocesstextwithcomputersino