본문

서브메뉴

Data Driven Simulation Frameworks for Operational Decision Making
Data Driven Simulation Frameworks for Operational Decision Making
Data Driven Simulation Frameworks for Operational Decision Making

Detailed Information

자료유형  
 학위논문 서양
최종처리일시  
20260202103105
ISBN  
9798288863073
DDC  
658
저자명  
Aykanat, Dilara.
서명/저자  
Data Driven Simulation Frameworks for Operational Decision Making
발행사항  
[Sl] : University of California, Berkeley, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
97 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-01, Section: B.
주기사항  
Advisor: Zheng, Zeyu.
학위논문주기  
Thesis (Ph.D.)--University of California, Berkeley, 2025.
초록/해제  
요약Many real world processes can be modeled as stochastic systems that necessitate simulations to evaluate multiple scenarios. However, uncertainties in input data, limited historical information, high computational costs and the need for extensive simulations create several challenges: i) The exact optimization of the final objective might be infeasible or costly which makes the simulation necessary. ii) The past data may not be sufficient to estimate the input parameters with high confidence. iii) Errors in the input distribution estimates can propagate and potentially affect the final objective predictions which may lead to suboptimal decisions. This dissertation presents simple and versatile simulation frameworks for data-driven operational decision making: i) We introduce a regularized simulation setting where the task's objective is considered in the training of the generators for a more accurate scenario planning. ii) We propose a simple general approach to the classical newsvendor problem with limited past information that balances the uncertainty risk. Our solutions can be adapted to any type of generator, predictor and do not rely on specific assumptions regarding the final objective function. Finally, we introduce a Taboo game benchmark for large language models to evaluate their ability to express words under constraints, which has potential applications in AI safety, training and content moderation.
일반주제명  
Industrial engineering
일반주제명  
Computer science
키워드  
Large language models
키워드  
Newsvendor problem
키워드  
Content moderation
기타저자  
University of California, Berkeley Industrial Engineering & Operations Research
기본자료저록  
Dissertations Abstracts International. 87-01B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017356943
■00520260202103105
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798288863073
■035    ▼a(MiAaPQ)AAI31935383
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a658
■1001  ▼aAykanat,  Dilara.
■24510▼aData  Driven  Simulation  Frameworks  for  Operational  Decision  Making
■260    ▼a[Sl]▼bUniversity  of  California,  Berkeley▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a97  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-01,  Section:  B.
■500    ▼aAdvisor:  Zheng,  Zeyu.
■5021  ▼aThesis  (Ph.D.)--University  of  California,  Berkeley,  2025.
■520    ▼aMany  real  world  processes  can  be  modeled  as  stochastic  systems  that  necessitate  simulations  to  evaluate  multiple  scenarios.  However,  uncertainties  in  input  data,  limited  historical  information,  high  computational  costs  and  the  need  for  extensive  simulations  create  several  challenges:  i)  The  exact  optimization  of  the  final  objective  might  be  infeasible  or  costly  which  makes  the  simulation  necessary.  ii)  The  past  data  may  not  be  sufficient  to  estimate  the  input  parameters  with  high  confidence.  iii)  Errors  in  the  input  distribution  estimates  can  propagate  and  potentially  affect  the  final  objective  predictions  which  may  lead  to  suboptimal  decisions.  This  dissertation  presents  simple  and  versatile  simulation  frameworks  for  data-driven  operational  decision  making:  i)  We  introduce  a  regularized  simulation  setting  where  the  task's  objective  is  considered  in  the  training  of  the  generators  for  a  more  accurate  scenario  planning.  ii)  We  propose  a  simple  general  approach  to  the  classical  newsvendor  problem  with  limited  past  information  that  balances  the  uncertainty  risk.  Our  solutions  can  be  adapted  to  any  type  of  generator,  predictor  and  do  not  rely  on  specific  assumptions  regarding  the  final  objective  function.  Finally,  we  introduce  a  Taboo  game  benchmark  for  large  language  models  to  evaluate  their  ability  to  express  words  under  constraints,  which  has  potential  applications  in  AI  safety,  training  and  content  moderation.
■590    ▼aSchool  code:  0028.
■650  4▼aIndustrial  engineering
■650  4▼aComputer  science
■653    ▼aLarge  language  models
■653    ▼aNewsvendor  problem
■653    ▼aContent  moderation
■690    ▼a0546
■690    ▼a0984
■690    ▼a0800
■71020▼aUniversity  of  California,  Berkeley▼bIndustrial  Engineering  &  Operations  Research.
■7730  ▼tDissertations  Abstracts  International▼g87-01B.
■790    ▼a0028
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17356943▼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
    TF14663 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.