본문

서브메뉴

Statistical Learning and Optimization Under Distribution Shift
Statistical Learning and Optimization Under Distribution Shift
Statistical Learning and Optimization Under Distribution Shift

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202104822
ISBN  
9798293820047
DDC  
310
저자명  
Zhao, Boxin.
서명/저자  
Statistical Learning and Optimization Under Distribution Shift
발행사항  
[Sl] : The University of Chicago, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
267 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-03, Section: B.
주기사항  
Advisor: Kolar, Mladen;Ma, Cong.
학위논문주기  
Thesis (Ph.D.)--The University of Chicago, 2025.
초록/해제  
요약Modern machine learning models are increasingly deployed in settings where the classical assumption of independent and identically distributed (i.i.d.) data is violated due to distribution shifts across domains, populations, or time. This dissertation develops new methodologies in statistical learning and optimization that address such shifts through the unifying principle of adaptation-distinguishing between components that remain stable (preserved) and those that vary (adapted) across environments.Chapter 2 introduces Trans-Glasso, a two-stage procedure for precision matrix estimation under distribution shift. The method exploits a shared sparsity pattern across domains as the preserved structure and adjusts for domain-specific deviations via differential network estimation. We establish non-asymptotic error bounds, prove minimax optimality, and demonstrate the method's practical effectiveness on both simulated and real-world data.Chapter 3 presents SMART, a spectral regularization framework for multi-task learning. Assuming that the singular subspaces of the regression matrix are preserved while the projection weights vary across tasks, SMART estimates the target model via a nonconvex optimization problem regularized by source-informed subspaces. The method is supported by theoretical guarantees and achieves strong empirical performance.Chapter 4 introduces an online client sampling method for federated optimization under data heterogeneity. It leverages the slowly varying informativeness of clients as the preserved structure and dynamically adapts the sampling distribution. Formulated as an online learning problem with bandit feedback, the algorithm builds on Online Stochastic Mirror Descent and achieves consistent improvements over uniform sampling in both theory and practice.Collectively, these contributions advance a unified framework for learning and optimization under distribution shift by systematically decomposing each problem into preserved and adapted components. This framework enables the principled design of algorithms that are both theoretically grounded and practically effective in the presence of heterogeneous data.
일반주제명  
Statistics
일반주제명  
Computer science
일반주제명  
Applied mathematics
키워드  
Distribution shifts
키워드  
Federated learning
키워드  
Graphical models
키워드  
Multi-task learning
키워드  
Statistical learning
키워드  
Transfer learning
기타저자  
The University of Chicago.
기본자료저록  
Dissertations Abstracts International. 87-03B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017359020
■00520260202104822
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798293820047
■035    ▼a(MiAaPQ)AAI32169442
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a310
■1001  ▼aZhao,  Boxin.▼0(orcid)0009-0005-8385-8313
■24510▼aStatistical  Learning  and  Optimization  Under  Distribution  Shift
■260    ▼a[Sl]▼bThe  University  of  Chicago▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a267  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-03,  Section:  B.
■500    ▼aAdvisor:  Kolar,  Mladen;Ma,  Cong.
■5021  ▼aThesis  (Ph.D.)--The  University  of  Chicago,  2025.
■520    ▼aModern  machine  learning  models  are  increasingly  deployed  in  settings  where  the  classical  assumption  of  independent  and  identically  distributed  (i.i.d.)  data  is  violated  due  to  distribution  shifts  across  domains,  populations,  or  time.  This  dissertation  develops  new  methodologies  in  statistical  learning  and  optimization  that  address  such  shifts  through  the  unifying  principle  of  adaptation-distinguishing  between  components  that  remain  stable  (preserved)  and  those  that  vary  (adapted)  across  environments.Chapter  2  introduces  Trans-Glasso,  a  two-stage  procedure  for  precision  matrix  estimation  under  distribution  shift.  The  method  exploits  a  shared  sparsity  pattern  across  domains  as  the  preserved  structure  and  adjusts  for  domain-specific  deviations  via  differential  network  estimation.  We  establish  non-asymptotic  error  bounds,  prove  minimax  optimality,  and  demonstrate  the  method's  practical  effectiveness  on  both  simulated  and  real-world  data.Chapter  3  presents  SMART,  a  spectral  regularization  framework  for  multi-task  learning.  Assuming  that  the  singular  subspaces  of  the  regression  matrix  are  preserved  while  the  projection  weights  vary  across  tasks,  SMART  estimates  the  target  model  via  a  nonconvex  optimization  problem  regularized  by  source-informed  subspaces.  The  method  is  supported  by  theoretical  guarantees  and  achieves  strong  empirical  performance.Chapter  4  introduces  an  online  client  sampling  method  for  federated  optimization  under  data  heterogeneity.  It  leverages  the  slowly  varying  informativeness  of  clients  as  the  preserved  structure  and  dynamically  adapts  the  sampling  distribution.  Formulated  as  an  online  learning  problem  with  bandit  feedback,  the  algorithm  builds  on  Online  Stochastic  Mirror  Descent  and  achieves  consistent  improvements  over  uniform  sampling  in  both  theory  and  practice.Collectively,  these  contributions  advance  a  unified  framework  for  learning  and  optimization  under  distribution  shift  by  systematically  decomposing  each  problem  into  preserved  and  adapted  components.  This  framework  enables  the  principled  design  of  algorithms  that  are  both  theoretically  grounded  and  practically  effective  in  the  presence  of  heterogeneous  data.
■590    ▼aSchool  code:  0330.
■650  4▼aStatistics
■650  4▼aComputer  science
■650  4▼aApplied  mathematics
■653    ▼aDistribution  shifts
■653    ▼aFederated  learning
■653    ▼aGraphical  models
■653    ▼aMulti-task  learning
■653    ▼aStatistical  learning
■653    ▼aTransfer  learning
■690    ▼a0463
■690    ▼a0984
■690    ▼a0364
■71020▼aThe  University  of  Chicago.
■7730  ▼tDissertations  Abstracts  International▼g87-03B.
■790    ▼a0330
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17359020▼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
    TF17203 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.