본문

서브메뉴

Large-Scale Optimization for Deep Neural Network Architecture: a Dynamical System Theory Perspective
Large-Scale Optimization for Deep Neural Network Architecture: a Dynamical System Theory P...
Large-Scale Optimization for Deep Neural Network Architecture: a Dynamical System Theory Perspective

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202105516
ISBN  
9798263341909
DDC  
519.7
저자명  
Liu, Guan-Horng.
서명/저자  
Large-Scale Optimization for Deep Neural Network Architecture: a Dynamical System Theory Perspective
발행사항  
[Sl] : Georgia Institute of Technology, 2024
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2024
형태사항  
173 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-05, Section: B.
주기사항  
Advisor: Theodorou, Evangelos A.
학위논문주기  
Thesis (Ph.D.)--Georgia Institute of Technology, 2024.
초록/해제  
요약Optimization of deep neural networks (DNNs) has been a driving force in the advance-ment of modern machine learning and artificial intelligence. Despite efforts to design DNN architectures that leverage domain-specific knowledge, the development of optimization al-gorithms for training these million-parameter functions has often progressed independently of architectural innovations. This thesis delves into large-scale optimization methods that are not only aware of but also leverage the underlying deep architectural structures being optimized. Specifically, we demonstrate that the dynamical system and optimal control the-ory pave a profound foundation for algorithmic characterization in this unexplored avenue.Optimal control, in its broadest sense, examines the principle of optimization over dy-namical systems. This methodological perspective naturally arises in training neural dif-ferential equations and can be applied to standard DNNs by interpreting layer propaga-tion as discrete timesteps along a dynamical system, with Backpropagation emerging as an approximate dynamic programming method. Through development, we emphasize the significance of control-theoretic components such as differential programming, nonlinear Feynman-Kac, and path integral theory, which unify existing optimization methods while extending them to handle a broader class of complex dynamics and problem setups that can otherwise be hard to adapt or foresee.Our work demonstrates the broad applicability of control-theoretic optimization meth-ods in learning various deep architectures, including convolutional networks, neural ordi-nary differential equations, and neural stochastic differential equations such as denoising diffusion models. The resulting computational frameworks improve test-time performance and inference efficiency, enhance training robustness against unstable hyperparameters, ac-celerate convergence in terms of wall-clock time, and are applicable to a wide range of scientific problems, including image generation, restoration, translation, as well as solving mean-field games and opinion modeling.
일반주제명  
Dynamic programming
일반주제명  
Back propagation
일반주제명  
Decision making
일반주제명  
Neural networks
일반주제명  
Diffusion models
일반주제명  
Dynamical systems
일반주제명  
Visualization
일반주제명  
Ordinary differential equations
일반주제명  
Games
일반주제명  
System theory
일반주제명  
Computer science
일반주제명  
Mathematics
기타저자  
Georgia Institute of Technology.
기본자료저록  
Dissertations Abstracts International. 87-05B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2024        us                              c    eng  d
■001000017360388
■00520260202105516
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798263341909
■035    ▼a(MiAaPQ)AAI32309352
■035    ▼a(MiAaPQ)GeorgiaTech75655
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a519.7
■1001  ▼aLiu,  Guan-Horng.
■24510▼aLarge-Scale  Optimization  for  Deep  Neural  Network  Architecture:  a  Dynamical  System  Theory  Perspective
■260    ▼a[Sl]▼bGeorgia  Institute  of  Technology▼c2024
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2024
■300    ▼a173  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-05,  Section:  B.
■500    ▼aAdvisor:  Theodorou,  Evangelos  A.
■5021  ▼aThesis  (Ph.D.)--Georgia  Institute  of  Technology,  2024.
■520    ▼aOptimization  of  deep  neural  networks  (DNNs)  has  been  a  driving  force  in  the  advance-ment  of  modern  machine  learning  and  artificial  intelligence.  Despite  efforts  to  design  DNN  architectures  that  leverage  domain-specific  knowledge,  the  development  of  optimization  al-gorithms  for  training  these  million-parameter  functions  has  often  progressed  independently  of  architectural  innovations.  This  thesis  delves  into  large-scale  optimization  methods  that  are  not  only  aware  of  but  also  leverage  the  underlying  deep  architectural  structures  being  optimized.  Specifically,  we  demonstrate  that  the  dynamical  system  and  optimal  control  the-ory  pave  a  profound  foundation  for  algorithmic  characterization  in  this  unexplored  avenue.Optimal  control,  in  its  broadest  sense,  examines  the  principle  of  optimization  over  dy-namical  systems.  This  methodological  perspective  naturally  arises  in  training  neural  dif-ferential  equations  and  can  be  applied  to  standard  DNNs  by  interpreting  layer  propaga-tion  as  discrete  timesteps  along  a  dynamical  system,  with  Backpropagation  emerging  as  an  approximate  dynamic  programming  method.  Through  development,  we  emphasize  the  significance  of  control-theoretic  components  such  as  differential  programming,  nonlinear  Feynman-Kac,  and  path  integral  theory,  which  unify  existing  optimization  methods  while  extending  them  to  handle  a  broader  class  of  complex  dynamics  and  problem  setups  that  can  otherwise  be  hard  to  adapt  or  foresee.Our  work  demonstrates  the  broad  applicability  of  control-theoretic  optimization  meth-ods  in  learning  various  deep  architectures,  including  convolutional  networks,  neural  ordi-nary  differential  equations,  and  neural  stochastic  differential  equations  such  as  denoising  diffusion  models.  The  resulting  computational  frameworks  improve  test-time  performance  and  inference  efficiency,  enhance  training  robustness  against  unstable  hyperparameters,  ac-celerate  convergence  in  terms  of  wall-clock  time,  and  are  applicable  to  a  wide  range  of  scientific  problems,  including  image  generation,  restoration,  translation,  as  well  as  solving  mean-field  games  and  opinion  modeling.
■590    ▼aSchool  code:  0078.
■650  4▼aDynamic  programming
■650  4▼aBack  propagation
■650  4▼aDecision  making
■650  4▼aNeural  networks
■650  4▼aDiffusion  models
■650  4▼aDynamical  systems
■650  4▼aVisualization
■650  4▼aOrdinary  differential  equations
■650  4▼aGames
■650  4▼aSystem  theory
■650  4▼aComputer  science
■650  4▼aMathematics
■690    ▼a0800
■690    ▼a0984
■690    ▼a0405
■71020▼aGeorgia  Institute  of  Technology.
■7730  ▼tDissertations  Abstracts  International▼g87-05B.
■790    ▼a0078
■791    ▼aPh.D.
■792    ▼a2024
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360388▼nKERIS▼z이  자료의  원문은  한국교육학술정보원에서  제공합니다.

Preview

Export

ChatGPT Discussion

AI Recommended Related Books


    New Books MORE
    Statistics for the past 3 years. Go to brief

    Подробнее информация.

    • Бронирование
    • не существует
    • моя папка
    • Первый запрос зрения
    • Non-Book Loan Application
    • Nighttime Book Loan Application
    материал
    Reg No. Количество платежных Местоположение статус Ленд информации
    TF17133 전자도서 대출가능 My Folder 부재도서신고 비도서대출신청 야간 도서대출신청

    * Бронирование доступны в заимствований книги. Чтобы сделать предварительный заказ, пожалуйста, нажмите кнопку бронирование

    Books borrowed together with this book

    Related Popular Books

    Available after logging in.