返回
convolutional neural networks for sentence

convolutional neural networks for sentence

ID：34104137

大小：672.36 KB

页数：34页

时间：2019-03-03

convolutional neural networks for sentence_第1页
convolutional neural networks for sentence_第2页
convolutional neural networks for sentence_第3页
convolutional neural networks for sentence_第4页
convolutional neural networks for sentence_第5页
资源描述：

《convolutional neural networks for sentence》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、ConvolutionalNeuralNetworksforSentenceClassi cationConvolutionalNeuralNetworksforSentenceClassi cationYoonKimNewYorkUniversity1/34ConvolutionalNeuralNetworksforSentenceClassi cationAgendaWordEmbeddingsClassi cationRecursiveNeuralTensorNetworksConvolutionalNeuralNetwor

2、ksExperimentsConclusion2/34ConvolutionalNeuralNetworksforSentenceClassi cationWordEmbeddingsDeeplearninginNaturalLanguageProcessingIDeeplearninghasachievedstate-of-the-artresultsincomputervision(Krizhevskyetal.,2012)andspeech(Gravesetal.,2013).INLP:fastbecoming(alread

3、yis)ahotareaofresearch.IMuchoftheworkinvolveslearningwordembeddingsandperformingcompositionoverthelearnedembeddingsforNLPtasks.3/34ConvolutionalNeuralNetworksforSentenceClassi cationWordEmbeddingsWordEmbeddings(orWordVectors)ITraditionalNLP:Wordsaretreatedasindices(or

4、one-hot"vectorsinRV)IEverywordisorthogonaltooneanother.Iwmotherwfather=0ICanweembedwordsinRDwithDVsuchthatsemanticallyclosewordsarelikewise`close'inRD?(i.e.wmotherwfather>0)IYes!IDon't(necessarily)needdeeplearningforthis:LatentSemanticAnalysis,LatentDirichletAlloc

5、ation,orsimplecontextcountsallgivedenserepresentations.4/34ConvolutionalNeuralNetworksforSentenceClassi cationWordEmbeddingsNeuralLanguageModels(NLM)IAnotherwaytoobtainwordembeddings.IWordsareprojectedfromRVtoRDviaahiddenlayer.IDisahyperparametertobetuned.IVariousarch

6、itecturesexist.Simpleonesarepopularthesedays(right).IVeryfast

7、cantrainonbillionsoftokensinonedaywithasinglemachine.Figure1:SkipgramarchitectureofMikolovetal.(2013)5/34ConvolutionalNeuralNetworksforSentenceClassi cationWordEmbeddingsLinguisticregularitiesintheobtainede

8、mbeddingsIThelearnedembeddingsencodesemanticandsyntacticregularities:IwbigwbiggerwslowwslowerIwfrancewpariswkoreawseoulIThesearecool,butnotnecessarilyuniquetoneurallanguagemodels.[...]theneuralembeddingprocessisnotdiscoveringnovelpatterns,butratherisdoingaremar

9、kablejobatpreservingthepatternsinherentintheword-contextco-occurrencematrix."LevyandGoldberg,LinguisticRegularitiesinSparse

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。