Transfer Learning with Dynamic Distribution Adaptation

doi:10.1145/3360309

	Transfer Learning with Dynamic Distribution Adaptation
	Wang, Jindong 1; Chen, Yiqiang 2; Feng, Wenjie 2; Yu, Han 3; Huang, Meiyu 4; Yang, Qiang 5
	2020-02-01
发表期刊	ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY
ISSN	2157-6904
卷号	11 期号:1 页码:25
摘要	Transfer learning aims to learn robust classifiers for the target domain by leveraging knowledge from a source domain. Since the source and the target domains are usually from different distributions, existing methods mainly focus on adapting the cross-domain marginal or conditional distributions. However, in real applications, the marginal and conditional distributions usually have different contributions to the domain discrepancy. Existing methods fail to quantitatively evaluate the different importance of these two distributions, which will result in unsatisfactory transfer performance. In this article, we propose a novel concept called Dynamic Distribution Adaptation (DDA), which is capable of quantitatively evaluating the relative importance of each distribution. DDA can be easily incorporated into the framework of structural risk minimization to solve transfer learning problems. On the basis of DDA, we propose two novel learning algorithms: (1) ManifoldDynamic DistributionAdaptation (MDDA) for traditional transfer learning, and (2) Dynamic Distribution Adaptation Network (DDAN) for deep transfer learning. Extensive experiments demonstrate that MDDA and DDAN significantly improve the transfer learning performance and set up a strong baseline over the latest deep and adversarial methods on digits recognition, sentiment analysis, and image classification. More importantly, it is shown that marginal and conditional distributions have different contributions to the domain divergence, and our DDA is able to provide good quantitative evaluation of their relative importance, which leads to better performance. We believe this observation can be helpful for future research in transfer learning.
关键词	Transfer learning domain adaptation distribution alignment deep learning subspace learning kernel method
DOI	10.1145/3360309
收录类别	SCI
语种	英语
资助项目	National Key R&D Program of China[2016YFB1001200] ; NSFC[61572471] ; NSFC[61972383] ; Hong Kong CERG projects[16209715] ; Hong Kong CERG projects[16244616] ; Nanyang Technological University, Nanyang Assistant Professorship (NAP) ; Beijing Municipal Science & Technology Commission[Z171100000117017]
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Information Systems
WOS记录号	WOS:000535726400006
出版者	ASSOC COMPUTING MACHINERY
引用统计	被引频次：221[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/15323
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Chen, Yiqiang
作者单位	1.Microsoft Res Asia, 5 Danling St, Beijing 100080, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, 6 Kexueyuan South Rd, Beijing 100190, Peoples R China 3.Nanyang Technol Univ, Sch Comp Sci & Engn, Singapore, Singapore 4.China Acad Space Technol, Qian Xuesen Lab Space Technol, Beijing, Peoples R China 5.Hong Kong Univ Sci & Technol, Dept Comp Sci & Engn, Kowloon, Hong Kong, Peoples R China
推荐引用方式 GB/T 7714	Wang, Jindong,Chen, Yiqiang,Feng, Wenjie,et al. Transfer Learning with Dynamic Distribution Adaptation[J]. ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY,2020,11(1):25.
APA	Wang, Jindong,Chen, Yiqiang,Feng, Wenjie,Yu, Han,Huang, Meiyu,&Yang, Qiang.(2020).Transfer Learning with Dynamic Distribution Adaptation.ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY,11(1),25.
MLA	Wang, Jindong,et al."Transfer Learning with Dynamic Distribution Adaptation".ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY 11.1(2020):25.