基于张量网络的多标签学习方法  

Multi-label learning method based on tensor network

作　　者：李丹萌 张晨光 刘天[1,2] 杜雪姣 Li Danmeng;Zhang Chenguang;Liu Tian;Du Xuejiao(School of Science,Hainan University,Haikou 570228,China;Hainan Key Laboratory of Engineering Modeling and Statistical Computing,Hainan University,Haikou 570228,China)

机构地区：[1]海南大学理学院,海南海口570228 [2]海南大学海南省工程建模与统计计算重点实验室,海南海口570228

出　　处：《海南大学学报（自然科学版）》2023年第4期335-342,共8页Natural Science Journal of Hainan University

基　　金：国家自然科学基金(62166016);海南省院士创新平台科研项目。

摘　　要：利用关系分类模型,将标签之间的相关性以及特征对标签相关性的影响形式化为分数模型,通过要求模型能够区分真实数据和噪声数据的得分建立了基于张量网络的多标签分类模型.多个数据集上的实验表明,相较于传统多标签学习方法和已有考察标签相关性的多标签学习方法,本文方法在平均精确度和错误率等多标签评价指标上提升近一倍,且拥有更低的计算成本.In the report,the relational classification model was used to formalize the correlation between labels and the effects of features on label correlation into a fractional model.A multi-label classification model based on tensor network was established by requiring the model to be able to distinguish the scores of real data and noise data.The results on multiple data sets showed that compared with traditional multi-label learning methods and existing multi-label learning methods to investigate label correlation,the method has nearly doubled the av-erage accuracy and error rate of multi-label evaluation indicators,and has a lower computational cost.

关 键 词：多标签学习 标签相关性 特征相关性 关系模型 张量网络 

分 类 号：N32[自然科学总论]