Non-negative matrix factorization based modeling and training algorithm for multi-label learning

作     者：Liang SUN Hongwei GE Wenjing KANG 

作者机构：College of Computer Science and TechnologyDalian University of TechnologyDalian 116024China 

出 版 物：《Frontiers of Computer Science》 (中国计算机科学前沿（英文版）)

年 卷 期：2019年第13卷第6期

页      面：1243-1254页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：support of the National Natural Science Foundation of China(Grant Nos.61402076,61572104,61103146) the Fundamental Research Funds for the Central Universities(DUT17JC04) the Project of the Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education,Jilin University(93K172017K03) 

主　　题：multi-label learning non-negative least square optimization non-negative matrix factorization smoothness assumption 

摘      要：Multi-label learning is more complicated than single-label learning since the semantics of the instances are usually overlapped and not *** effectiveness of many algorithms often fails when the correlations in the feature and label space are not fully *** this end,we propose a novel non-negative matrix factorization(NMF)based modeling and training algorithm that learns from both the adjacencies of the instances and the labels of the training *** the modeling process,a set of generators are constructed,and the associations among generators,instances,and labels are set up,with which the label prediction is *** the training process,the parameters involved in the process of modeling are ***,an NMF based algorithm is proposed to determine the associations between generators and instances,and a non-negative least square optimization algorithm is applied to determine the associations between generators and *** proposed algorithm fully takes the advantage of smoothness assumption,so that the labels are properly *** experiments were carried out on six set of *** results demonstrate the effectiveness of the proposed algorithms.