典型文献
Label correlation for partial label learning
文献摘要：
Partial label learning aims to learn a multi-class classi-fier,where each training example corresponds to a set of candi-date labels among which only one is correct.Most studies in the label space have only focused on the difference between candi-date labels and non-candidate labels.So far,however,there has been little discussion about the label correlation in the partial label learning.This paper begins with a research on the label correlation,followed by the establishment of a unified frame-work that integrates the label correlation,the adaptive graph,and the semantic difference maximization criterion.This work generates fresh insight into the acquisition of the learning infor-mation from the label space.Specifically,the label correlation is calculated from the candidate label set and is utilized to obtain the similarity of each pair of instances in the label space.After that,the labeling confidence for each instance is updated by the smoothness assumption that two instances should be similar outputs in the label space if they are close in the feature space.At last,an effective optimization program is utilized to solve the unified framework.Extensive experiments on artificial and real-world data sets indicate the superiority of our proposed method to state-of-art partial label learning methods.
文献关键词：
中图分类号：
作者姓名：
GE Lingchi;FANG Min;LI Haikun;CHEN Bo
作者机构：
School of Computer Science and Technology,Xidian University,Xi'an 710071,China
文献出处：
引用格式：
[1]GE Lingchi;FANG Min;LI Haikun;CHEN Bo-.Label correlation for partial label learning)[J].系统工程与电子技术（英文版）,2022(05):1043-1051
A类：
B类：
Label,correlation,partial,learning,Partial,aims,multi,classi,fier,where,each,training,example,corresponds,labels,among,which,only,one,correct,Most,studies,space,have,focused,difference,between,candidate,So,far,however,there,has,been,little,discussion,about,This,paper,begins,research,followed,by,establishment,unified,that,integrates,adaptive,graph,semantic,maximization,criterion,generates,fresh,insight,into,acquisition,infor,mation,from,Specifically,calculated,utilized,obtain,similarity,pair,instances,After,labeling,confidence,updated,smoothness,assumption,two,should,outputs,they,are,close,feature,At,last,effective,optimization,program,solve,framework,Extensive,experiments,artificial,real,world,data,sets,indicate,superiority,our,proposed,state,methods
AB值：
0.535966
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。
