Certifiable Out-of-Distribution Generalization

	Certifiable Out-of-Distribution Generalization
	Ye, Nanyang 1; Zhu, Lin 1; Wang, Jia 2; Zeng, Zhaoyu 1; Shao, Jiayao 3; Peng, Chensheng 1; Pan, Bikang4 ; Li, Kaican 5; Zhu, Jun 6
	2023-06-27
会议录名称	PROCEEDINGS OF THE 37TH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, AAAI 2023
ISSN	2159-5399
卷号	37
页码	10927-10935
发表状态	已发表
摘要	Machine learning methods suffer from test-time performance degeneration when faced with out-of-distribution (OoD) data whose distribution is not necessarily the same as training data distribution. Although a plethora of algorithms have been proposed to mitigate this issue, it has been demonstrated that achieving better performance than ERM simultaneously on different types of distributional shift datasets is challenging for existing approaches. Besides, it is unknown how and to what extent these methods work on any OoD datum without theoretical guarantees. In this paper, we propose a certifiable out-of-distribution generalization method that provides provable OoD generalization performance guarantees via a functional optimization framework leveraging random distributions and max-margin learning for each input datum. With this approach, the proposed algorithmic scheme can provide certified accuracy for each input datum’s prediction on the semantic space and achieves better performance simultaneously on OoD datasets dominated by correlation shifts or diversity shifts. Our code is available at https://github.com/ZlatanWilliams/StochasticDisturbanceLearning. Copyright © 2023, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.
会议录编者/会议主办者	Association for the Advancement of Artificial Intelligence
关键词	Artificial intelligence Learning systems Data distribution Generalisation Generalization performance Input datas Machine learning methods Performance Performance guarantees Test time Theoretical guarantees Training data
会议名称	37th AAAI Conference on Artificial Intelligence, AAAI 2023
出版地	2275 E BAYSHORE RD, STE 160, PALO ALTO, CA 94303 USA
会议地点	Washington, DC, United states
会议日期	February 7, 2023 - February 14, 2023
URL	查看原文
收录类别	EI ; CPCI-S
语种	英语
资助项目	National Natural Science Foundation of China[62106139]
WOS研究方向	Computer Science ; Mathematics
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Theory & Methods ; Mathematics, Applied
WOS记录号	WOS:001243747800070
出版者	AAAI Press
EI入藏号	20233414600908
EI主题词	Semantics
EISSN	2374-3468
EI分类号	723.4 Artificial Intelligence
原始文献类型	Conference article (CA)
引用统计	正在获取...
文献类型	会议论文
条目标识符	https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/348714
专题	信息科学与技术学院_博士生
通讯作者	Ye, Nanyang
作者单位	1.Shanghai Jiao Tong University, Shanghai, China 2.University of Cambridge, Cambridge, United Kingdom 3.University of Warwick, Warwick, United Kingdom 4.ShanghaiTech University, Shanghai, China 5.Huawei Noah’s Ark Lab., Hong Kong 6.Tsinghua University, Beijing, China
推荐引用方式 GB/T 7714	Ye, Nanyang,Zhu, Lin,Wang, Jia,et al. Certifiable Out-of-Distribution Generalization[C]//Association for the Advancement of Artificial Intelligence. 2275 E BAYSHORE RD, STE 160, PALO ALTO, CA 94303 USA:AAAI Press,2023:10927-10935.