본문

서브메뉴

Certifiably Trustworthy Deep Learning Systems at Scale
Certifiably Trustworthy Deep Learning Systems at Scale
Certifiably Trustworthy Deep Learning Systems at Scale

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260209102856
ISBN  
9798291574300
DDC  
004
저자명  
Li, Linyi.
서명/저자  
Certifiably Trustworthy Deep Learning Systems at Scale
발행사항  
[Sl] : University of Illinois at Urbana-Champaign, 2023
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2023
형태사항  
547 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-02, Section: B.
주기사항  
Advisor: Li, Bo;Xie, Tao.
학위논문주기  
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2023.
초록/해제  
요약Great advances in deep learning (DL) have led to state-of-the-art performance on a wide range of challenging tasks. However, along with the rapid deployment of DL systems, several trustworthy threats arise, such as weak robustness against stealthy noise perturbations and natural transformations, bias across different subgroups, and lack of numerical reliability. These trustworthy threats have raised great concerns, especially when deploying DL systems in safety-critical scenarios such as autonomous driving and facial recognition for safeguarding. To defend against these common trustworthy threats, this thesis systematically proposes or enhances certification approaches and certified training approaches for DL systems, especially for large-scale DL systems.A certification approach can guarantee some properties of the DL system under some trustworthiness properties. For instance, the robustness certification approach can guarantee the worst-case test accuracy when the attacker imposes any input perturbations or transformations within some bounded range. A certified training approach can improve the DL system's guaranteed trustworthiness under a certain property by training the DL model, e.g., improving the guaranteed test accuracy above.This thesis begins with a systematic taxonomy of certification and certified training approaches. Then for several critical trustworthiness properties, this thesis proposes the corresponding certification and certified training approaches that lead to state-of-the-art tightness and scalability. These approaches are motivated by a few core principles, including dual problem analysis for randomized smoothing, general cutting planes for bound propagation, stratified sampling, subpopulation decomposition, and abstract interpretation. The effectiveness of the proposed approaches is supported by both theoretical analyses and empirical evaluations. The thesis is concluded with a discussion of limitations, challenges, and future directions towards achieving fully certifiable, reliable, and scalable machine learning.In summary, this thesis enables certification of various trustworthy properties for DL systems up to millions of parameters, representing a major step in certified deep learning, an important research topic in machine learning, computer security, and software engineering.
일반주제명  
Computer science
일반주제명  
Information technology
일반주제명  
Robotics
키워드  
Deep learning
키워드  
Certification
키워드  
Verification
키워드  
Trustworthy machine learning
키워드  
Machine learning security
기타저자  
University of Illinois at Urbana-Champaign Computer Science
기본자료저록  
Dissertations Abstracts International. 87-02B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260203s2023        us                              c    eng  d
■001000017365923
■00520260209102856
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798291574300
■035    ▼a(MiAaPQ)AAI32272137
■035    ▼a(MiAaPQ)httphdlhandlenet2142121953
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a004
■1001  ▼aLi,  Linyi.
■24510▼aCertifiably  Trustworthy  Deep  Learning  Systems  at  Scale
■260    ▼a[Sl]▼bUniversity  of  Illinois  at  Urbana-Champaign▼c2023
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2023
■300    ▼a547  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-02,  Section:  B.
■500    ▼aAdvisor:  Li,  Bo;Xie,  Tao.
■5021  ▼aThesis  (Ph.D.)--University  of  Illinois  at  Urbana-Champaign,  2023.
■520    ▼aGreat  advances  in  deep  learning  (DL)  have  led  to  state-of-the-art  performance  on  a  wide  range  of  challenging  tasks.  However,  along  with  the  rapid  deployment  of  DL  systems,  several  trustworthy  threats  arise,  such  as  weak  robustness  against  stealthy  noise  perturbations  and  natural  transformations,  bias  across  different  subgroups,  and  lack  of  numerical  reliability.  These  trustworthy  threats  have  raised  great  concerns,  especially  when  deploying  DL  systems  in  safety-critical  scenarios  such  as  autonomous  driving  and  facial  recognition  for  safeguarding.  To  defend  against  these  common  trustworthy  threats,  this  thesis  systematically  proposes  or  enhances  certification  approaches  and  certified  training  approaches  for  DL  systems,  especially  for  large-scale  DL  systems.A  certification  approach  can  guarantee  some  properties  of  the  DL  system  under  some  trustworthiness  properties.  For  instance,  the  robustness  certification  approach  can  guarantee  the  worst-case  test  accuracy  when  the  attacker  imposes  any  input  perturbations  or  transformations  within  some  bounded  range.  A  certified  training  approach  can  improve  the  DL  system's  guaranteed  trustworthiness  under  a  certain  property  by  training  the  DL  model,  e.g.,  improving  the  guaranteed  test  accuracy  above.This  thesis  begins  with  a  systematic  taxonomy  of  certification  and  certified  training  approaches.  Then  for  several  critical  trustworthiness  properties,  this  thesis  proposes  the  corresponding  certification  and  certified  training  approaches  that  lead  to  state-of-the-art  tightness  and  scalability.  These  approaches  are  motivated  by  a  few  core  principles,  including  dual  problem  analysis  for  randomized  smoothing,  general  cutting  planes  for  bound  propagation,  stratified  sampling,  subpopulation  decomposition,  and  abstract  interpretation.  The  effectiveness  of  the  proposed  approaches  is  supported  by  both  theoretical  analyses  and  empirical  evaluations.  The  thesis  is  concluded  with  a  discussion  of  limitations,  challenges,  and  future  directions  towards  achieving  fully  certifiable,  reliable,  and  scalable  machine  learning.In  summary,  this  thesis  enables  certification  of  various  trustworthy  properties  for  DL  systems  up  to  millions  of  parameters,  representing  a  major  step  in  certified  deep  learning,  an  important  research  topic  in  machine  learning,  computer  security,  and  software  engineering.
■590    ▼aSchool  code:  0090.
■650  4▼aComputer  science
■650  4▼aInformation  technology
■650  4▼aRobotics
■653    ▼aDeep  learning
■653    ▼aCertification
■653    ▼aVerification
■653    ▼aTrustworthy  machine  learning
■653    ▼aMachine  learning  security
■690    ▼a0984
■690    ▼a0800
■690    ▼a0489
■690    ▼a0771
■71020▼aUniversity  of  Illinois  at  Urbana-Champaign▼bComputer  Science.
■7730  ▼tDissertations  Abstracts  International▼g87-02B.
■790    ▼a0090
■791    ▼aPh.D.
■792    ▼a2023
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17365923▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

미리보기

내보내기

chatGPT토론

Ai 추천 관련 도서


    신착도서 더보기
    최근 3년간 통계입니다.

    소장정보

    • 예약
    • 소재불명신고
    • 나의폴더
    • 우선정리요청
    • 비도서대출신청
    • 야간 도서대출신청
    소장자료
    등록번호 청구기호 소장처 대출가능여부 대출정보
    TF15723 전자도서 대출가능 마이폴더 부재도서신고 비도서대출신청 야간 도서대출신청

    * 대출중인 자료에 한하여 예약이 가능합니다. 예약을 원하시면 예약버튼을 클릭하십시오.

    해당 도서를 다른 이용자가 함께 대출한 도서

    관련 인기도서

    로그인 후 이용 가능합니다.