典型文献
Smoother manifold for graph meta-learning
文献摘要：
Meta-learning provides a framework for the possibility of mimicking artificial intelligence. How-ever, data distribution of the training set fails to be consistent with the one of the testing set as the limited domain differences among them. These factors often result in poor generalization in existing meta-learning models. In this work, a novel smoother manifold for graph meta-learning ( SGML) is proposed, which derives the similarity parameters of node features from the relationship between nodes and edges in the graph structure, and then utilizes the similarity parameters to yield smoother manifold through embedded propagation module. Smoother manifold can naturally filter out noise from the most important components when generalizing the local mapping relationship to the global. Besides suiting for generalizing on unseen low data issues, the framework is capable to easily perform transductive inference. Experimental results on MiniImageNet and TieredImageNet consistently show that applying SGML to supervised and semi-supervised classification can improve the performance in reducing the noise of domain shift representation.
文献关键词：
中图分类号：
作者姓名：
ZHAO Wencang;WANG Chunxin;XU Changkai
作者机构：
College of Automation and Electronic Engineering,Qingdao University of Science and Technology,Qingdao 266061,P.R.China
文献出处：
引用格式：
[1]ZHAO Wencang;WANG Chunxin;XU Changkai-.Smoother manifold for graph meta-learning)[J].高技术通讯（英文版）,2022(01):48-55
A类：
Smoother,SGML,TieredImageNet
B类：
manifold,graph,meta,learning,provides,framework,possibility,mimicking,artificial,intelligence,How,ever,data,distribution,training,set,fails,testing,limited,domain,differences,among,them,These,factors,often,poor,generalization,existing,models,In,this,novel,smoother,proposed,which,derives,similarity,parameters,features,from,relationship,between,nodes,edges,structure,then,utilizes,yield,through,embedded,propagation,module,can,naturally,filter,out,noise,most,important,components,when,generalizing,local,mapping,global,Besides,suiting,unseen,low,issues,capable,easily,transductive,inference,Experimental,results,MiniImageNet,consistently,show,that,applying,supervised,semi,classification,improve,performance,reducing,shift,representation
AB值：
0.593849
相似文献
机标中图分类号，由域田数据科技根据网络公开资料自动分析生成，仅供学习研究参考。
