서브메뉴
검색
Principled Algorithms for Domain Adaptation and Generalization- [electronic resource]
Principled Algorithms for Domain Adaptation and Generalization- [electronic resource]
상세정보
- 자료유형
- 학위논문
- Control Number
- 0016934563
- International Standard Book Number
- 9798380485050
- Dewey Decimal Classification Number
- 004
- Main Entry-Personal Name
- Chen, Yining.
- Publication, Distribution, etc. (Imprint
- [S.l.] : Stanford University., 2023
- Publication, Distribution, etc. (Imprint
- Ann Arbor : ProQuest Dissertations & Theses, 2023
- Physical Description
- 1 online resource(229 p.)
- General Note
- Source: Dissertations Abstracts International, Volume: 85-04, Section: B.
- General Note
- Advisor: Ma, Tengyu.
- Dissertation Note
- Thesis (Ph.D.)--Stanford University, 2023.
- Restrictions on Access Note
- This item must not be sold to any third party vendors.
- Summary, Etc.
- 요약Machine learning models are increasingly applied to datasets different from the training datasets. The performance of models often degrades when tested on unseen scenarios. Empirically, many algorithms have been used for domain adaptation and generalization, but few methods have been able to surpass empirical risk minimization consistently on common benchmarks. Theoretically, traditional learning theory offers limited insights for distributional shift problems. The main goal of this thesis is to bridge the gap between the theory and practice for domain shift problems, and to develop principled algorithms that have better robustness guarantees.We study three domain shift problems with increased supervised from the target domain. We first study domain generalization where no target data is available during training. We show that feature-matching algorithms generalize better when the distinguishing property of the signal feature is indeed conditional distributional invariance. Next, we study domain adaptation where unlabeled target data is available. We show that self-training helps when the target is more diverse than the source. Lastly, we study active online learning under domain shift. We show that uncertainty sampling leads to better query-regret tradeoff when there is hidden domain structure. In all three problems, the synergy of explicit bias from the algorithm and implicit bias from the domain shift structure contributes to successful transfer between domains.
- Subject Added Entry-Topical Term
- Active learning.
- Subject Added Entry-Topical Term
- Computer science.
- Added Entry-Corporate Name
- Stanford University.
- Host Item Entry
- Dissertations Abstracts International. 85-04B.
- Host Item Entry
- Dissertation Abstract International
- Electronic Location and Access
- 로그인을 한후 보실 수 있는 자료입니다.
- Control Number
- joongbu:639850
MARC
008240219s2023 ulk 00 kor■001000016934563
■00520240214101626
■006m o d
■007cr#unu||||||||
■020 ▼a9798380485050
■035 ▼a(MiAaPQ)AAI30615214
■035 ▼a(MiAaPQ)STANFORDdm107nr6713
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a004
■1001 ▼aChen, Yining.
■24510▼aPrincipled Algorithms for Domain Adaptation and Generalization▼h[electronic resource]
■260 ▼a[S.l.]▼bStanford University. ▼c2023
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2023
■300 ▼a1 online resource(229 p.)
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-04, Section: B.
■500 ▼aAdvisor: Ma, Tengyu.
■5021 ▼aThesis (Ph.D.)--Stanford University, 2023.
■506 ▼aThis item must not be sold to any third party vendors.
■520 ▼aMachine learning models are increasingly applied to datasets different from the training datasets. The performance of models often degrades when tested on unseen scenarios. Empirically, many algorithms have been used for domain adaptation and generalization, but few methods have been able to surpass empirical risk minimization consistently on common benchmarks. Theoretically, traditional learning theory offers limited insights for distributional shift problems. The main goal of this thesis is to bridge the gap between the theory and practice for domain shift problems, and to develop principled algorithms that have better robustness guarantees.We study three domain shift problems with increased supervised from the target domain. We first study domain generalization where no target data is available during training. We show that feature-matching algorithms generalize better when the distinguishing property of the signal feature is indeed conditional distributional invariance. Next, we study domain adaptation where unlabeled target data is available. We show that self-training helps when the target is more diverse than the source. Lastly, we study active online learning under domain shift. We show that uncertainty sampling leads to better query-regret tradeoff when there is hidden domain structure. In all three problems, the synergy of explicit bias from the algorithm and implicit bias from the domain shift structure contributes to successful transfer between domains.
■590 ▼aSchool code: 0212.
■650 4▼aActive learning.
■650 4▼aComputer science.
■690 ▼a0800
■690 ▼a0984
■71020▼aStanford University.
■7730 ▼tDissertations Abstracts International▼g85-04B.
■773 ▼tDissertation Abstract International
■790 ▼a0212
■791 ▼aPh.D.
■792 ▼a2023
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T16934563▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
■980 ▼a202402▼f2024
미리보기
내보내기
chatGPT토론
Ai 추천 관련 도서
Info Détail de la recherche.
- Réservation
- 캠퍼스간 도서대출
- 서가에 없는 책 신고
- My Folder