OPM2L: An optimal instance partition-based multi-metric learning method for heterogeneous dataset classification
文献类型:期刊论文
作者 | Deng, Huiyuan1; Meng, Xiangzhu2; Wang, Huibing3; Feng, Lin1 |
刊名 | INFORMATION SCIENCES |
出版日期 | 2023-11-01 |
卷号 | 648页码:18 |
ISSN号 | 0020-0255 |
关键词 | Multi-metric learning Alternating direction method Nearest-neighbor classification Riemannian manifold |
DOI | 10.1016/j.ins.2023.119550 |
通讯作者 | Feng, Lin(fenglin@dlut.edu.cn) |
英文摘要 | Multi-metric learning-a method to learn multiple local metrics to reveal the feature's correlations of samples from different local regions-has become an essential tool to measure the similarities between instances from heterogeneous datasets. However, most existing cluster-based MML methods first partition the training data with a predefined metric and then learn multiple metrics via the local instances, leading to these two independent procedures fail to cooperate with each other. In this paper, we propose an Optimal instance Partition-based Multi-Metric Learning (OPM2L) method for heterogeneous dataset classification by unifying the instance partition and multiple local metrics learning into a single objective. In particular, multiple anchor centers together with a global metric are employed to assist the instance partition process. During the training, the shared information contained in local metrics is aggregated into the global metric by a dedicated regularizer, which improves the instance partition process and offers the subsequent multiple local metrics learning with more informative instances. Moreover, an efficient alternating direction technology is employed to seek a feasible solution to the proposed method. We further confirmed that the sub-problems can be settled with closed-form solutions, while the superiority of the proposed method is also proved by experimental results on extensive datasets. |
资助项目 | National Natural Science Foundation of PR China[61972064] ; Liaoning Revitalization Talents Program[XLYC1806006] ; Fundamental Research Funds for the Central Universities[DUT19RC (3) 012] |
WOS研究方向 | Computer Science |
语种 | 英语 |
出版者 | ELSEVIER SCIENCE INC |
WOS记录号 | WOS:001070036100001 |
资助机构 | National Natural Science Foundation of PR China ; Liaoning Revitalization Talents Program ; Fundamental Research Funds for the Central Universities |
源URL | [http://ir.ia.ac.cn/handle/173211/53110] |
专题 | 自动化研究所_智能感知与计算研究中心 |
通讯作者 | Feng, Lin |
作者单位 | 1.Dalian Univ Technol, Sch Innovat & Entrepreneurship, Dalian, Peoples R China 2.Chinese Acad Sci, Ctr Res Intelligent Percept & Comp, Inst Automat, Beijing, Peoples R China 3.Dalian Maritime Univ, Informat Sci & Technol Coll, Dalian 116024, Peoples R China |
推荐引用方式 GB/T 7714 | Deng, Huiyuan,Meng, Xiangzhu,Wang, Huibing,et al. OPM2L: An optimal instance partition-based multi-metric learning method for heterogeneous dataset classification[J]. INFORMATION SCIENCES,2023,648:18. |
APA | Deng, Huiyuan,Meng, Xiangzhu,Wang, Huibing,&Feng, Lin.(2023).OPM2L: An optimal instance partition-based multi-metric learning method for heterogeneous dataset classification.INFORMATION SCIENCES,648,18. |
MLA | Deng, Huiyuan,et al."OPM2L: An optimal instance partition-based multi-metric learning method for heterogeneous dataset classification".INFORMATION SCIENCES 648(2023):18. |
入库方式: OAI收割
来源:自动化研究所
浏览0
下载0
收藏0
其他版本
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。