본문

서브메뉴

Reinforcement Learning via Differentiable Simulation: Applications in Operations Management
Reinforcement Learning via Differentiable Simulation: Applications in Operations Managemen...
Reinforcement Learning via Differentiable Simulation: Applications in Operations Management

상세정보

자료유형  
 학위논문 서양
최종처리일시  
20260202104708
ISBN  
9798286498451
DDC  
658
저자명  
Che, Ethan.
서명/저자  
Reinforcement Learning via Differentiable Simulation: Applications in Operations Management
발행사항  
[Sl] : Columbia University, 2025
발행사항  
Ann Arbor : ProQuest Dissertations & Theses, 2025
형태사항  
348 p
주기사항  
Source: Dissertations Abstracts International, Volume: 87-01, Section: B.
주기사항  
Advisor: Dong, Jing;Namkoong, Hongseok.
학위논문주기  
Thesis (Ph.D.)--Columbia University, 2025.
초록/해제  
요약Reinforcement learning (RL) is a promising approach for solving dynamic decision-making problems that arise in operations management applications, including scheduling, routing, supply chain management, and adaptive experimentation. However, so far RL algorithms have not experienced widespread industrial adoption due to high sample complexity and difficulty in scaling to large-scale problems. To address these issues, this dissertation proposes a model-based approach to RL via differentiable simulation. Differentiable simulation is a technique for computing gradients of a system's dynamics with respect to any parameters or inputs, which has been widely used in physics and robotics, with roots in the stochastic simulation literature dating back to the 1980s-1990s with methodologies such as infinitesimal perturbation analysis (IPA) and generalized likelihood ratio estimation. We take inspiration from these ideas, which have typically been applied to systems with a handful of parameters (e.g. service rates or base-stock levels), and develop frameworks for applying them to reinforcement learning problems in different operations applications.The first study develops a differentiable optimization framework for batched adaptive experimentation. Sampling with large batches is a ubiquitious practice in online experimentation platforms, due to operational constraints as well as concerns over delayed feedback. Using large batch statistical approximations, we develop a differentiable formulation of Bayesian regret for adaptive experimentation which can be directly optimized, in contrast with existing heuristics that optimize proxies of regret. By using a Bayesain model of how additional data affects future posterior beliefs, this explicitly captures the trade-off between exploration and exploitation. Using differentiability of Bayesian regret, we propose a model-predictive control algorithm for adaptive experimentation in batched settings with short time horizons. The algorithm selects the sampling allocations-the number of samples assigned to each treatment arm-to directly minimize Bayesian regret and re-optimizes this after each batch. We find through extensive numerical experiments that our algorithm outperforms existing heuristics (e.g. Thompson Sampling) especially under short time horizons. We also make theoretical connections between model-predictive control and posterior sampling. The second study follows up on the first study and provides a generalization of our framework for contextual settings, multiple-objectives, constraints, and non-stationarity, which are common considerations in online experimentation platforms. We find that under these different considerations, the design of effective adaptive algorithms is highly non-trivial, as even simple strategies like uniform allocation can outperform adaptive algorithms like Thompson Sampling. In contrast, our framework allows the experimenter to flexibly include these considerations in the optimization problem, rather than developing bespoke algorithms for different situations and combinations of considerations. Not only does this optimization-based approach come with new theoretical guarantees in wide range of settings, in several experiments involving non-stationarity, constraints, and multiple-objectives, our algorithm is able to outperform existing heuristics such as Thompson Sampling.The third study develops a model-based reinforcement learning approach for queueing network control through differentiable discrete-event simulation. Model-free RL methods have been applied to queueing networks, but they are sample inefficient and suffer from slow convergence. Despite non-differentiability of the dynamics of queueing networks, we formulate a differentiable approximation of the dynamics and use this to obtain pathwise policy gradients that are orders of magnitude less noisy than standard REINFORCE gradients. Using these pathwise policy gradients, we are able to learn scheduling policies that outperform or match the policies trained with model-free RL with significantly less data. Unlike prior tailored approaches to queueing, our methods can flexibly handle realistic scenarios, including systems operating in non-stationary environments and those with non-exponential interarrival/service times.The fourth chapter is a theoretical study of the convergence of stochastic gradient descent with Markovian feedback. While there are many works on the convergence of policy gradient with REINFORCE-based gradient estimation, in this chapter we provide a convergence analysis for broader gradient estimators, including differentiable simulation or finite-difference based estimators. The core challenge is that the policy changes the environment and thereby affects the data used to update the policy. The influence of previous decisions on the data generated introduces bias in the gradient estimate, which presents a potential source of instability for online learning not present in the iid case. In this paper, we introduce simple criteria for the adaptively generated data stream to guarantee the convergence of SGD, which depend on the mixing time of the policy-induced dynamics. Our Lyapunov-function analysis allows one to translate existing stability analysis of stochastic systems studied in operations research into convergence rates for SGD, and we demonstrate this for queueing and inventory management problems. We also showcase how our result can be applied to study the sample complexity of an actor-critic policy gradient algorithm.
일반주제명  
Industrial engineering
일반주제명  
Statistics
키워드  
Adaptive experimentation
키워드  
Machine learning
키워드  
Queuing systems
키워드  
Reinforcement learning
키워드  
Infinitesimal perturbation analysis
기타저자  
Columbia University Operations Research
기본자료저록  
Dissertations Abstracts International. 87-01B.
전자적 위치 및 접속  
로그인 후 원문을 볼 수 있습니다.

MARC

 008260126s2025        us                              c    eng  d
■001000017358479
■00520260202104708
■006m          o    d                
■007cr#unu||||||||
■020    ▼a9798286498451
■035    ▼a(MiAaPQ)AAI32117746
■040    ▼aMiAaPQ▼cMiAaPQ
■0820  ▼a658
■1001  ▼aChe,  Ethan.
■24510▼aReinforcement  Learning  via  Differentiable  Simulation:  Applications  in  Operations  Management
■260    ▼a[Sl]▼bColumbia  University▼c2025
■260  1▼aAnn  Arbor▼bProQuest  Dissertations  &  Theses▼c2025
■300    ▼a348  p
■500    ▼aSource:  Dissertations  Abstracts  International,  Volume:  87-01,  Section:  B.
■500    ▼aAdvisor:  Dong,  Jing;Namkoong,  Hongseok.
■5021  ▼aThesis  (Ph.D.)--Columbia  University,  2025.
■520    ▼aReinforcement  learning  (RL)  is  a  promising  approach  for  solving  dynamic  decision-making  problems  that  arise  in  operations  management  applications,  including  scheduling,  routing,  supply  chain  management,  and  adaptive  experimentation.  However,  so  far  RL  algorithms  have  not  experienced  widespread  industrial  adoption  due  to  high  sample  complexity  and  difficulty  in  scaling  to  large-scale  problems.  To  address  these  issues,  this  dissertation  proposes  a  model-based  approach  to  RL  via  differentiable  simulation.  Differentiable  simulation  is  a  technique  for  computing  gradients  of  a  system's  dynamics  with  respect  to  any  parameters  or  inputs,  which  has  been  widely  used  in  physics  and  robotics,  with  roots  in  the  stochastic  simulation  literature  dating  back  to  the  1980s-1990s  with  methodologies  such  as  infinitesimal  perturbation  analysis  (IPA)  and  generalized  likelihood  ratio  estimation.  We  take  inspiration  from  these  ideas,  which  have  typically  been  applied  to  systems  with  a  handful  of  parameters  (e.g.  service  rates  or  base-stock  levels),  and  develop  frameworks  for  applying  them  to  reinforcement  learning  problems  in  different  operations  applications.The  first  study  develops  a  differentiable  optimization  framework  for  batched  adaptive  experimentation.  Sampling  with  large  batches  is  a  ubiquitious  practice  in  online  experimentation  platforms,  due  to  operational  constraints  as  well  as  concerns  over  delayed  feedback.  Using  large  batch  statistical  approximations,  we  develop  a  differentiable  formulation  of  Bayesian  regret  for  adaptive  experimentation  which  can  be  directly  optimized,  in  contrast  with  existing  heuristics  that  optimize  proxies  of  regret.  By  using  a  Bayesain  model  of  how  additional  data  affects  future  posterior  beliefs,  this  explicitly  captures  the  trade-off  between  exploration  and  exploitation.  Using  differentiability  of  Bayesian  regret,  we  propose  a  model-predictive  control  algorithm  for  adaptive  experimentation  in  batched  settings  with  short  time  horizons.  The  algorithm  selects  the  sampling  allocations-the  number  of  samples  assigned  to  each  treatment  arm-to  directly  minimize  Bayesian  regret  and  re-optimizes  this  after  each  batch.  We  find  through  extensive  numerical  experiments  that  our  algorithm  outperforms  existing  heuristics  (e.g.  Thompson  Sampling)  especially  under  short  time  horizons.  We  also  make  theoretical  connections  between  model-predictive  control  and  posterior  sampling. The  second  study  follows  up  on  the  first  study  and  provides  a  generalization  of  our  framework  for  contextual  settings,  multiple-objectives,  constraints,  and  non-stationarity,  which  are  common  considerations  in  online  experimentation  platforms.  We  find  that  under  these  different  considerations,  the  design  of  effective  adaptive  algorithms  is  highly  non-trivial,  as  even  simple  strategies  like  uniform  allocation  can  outperform  adaptive  algorithms  like  Thompson  Sampling.  In  contrast,  our  framework  allows  the  experimenter  to  flexibly  include  these  considerations  in  the  optimization  problem,  rather  than  developing  bespoke  algorithms  for  different  situations  and  combinations  of  considerations.  Not  only  does  this  optimization-based  approach  come  with  new  theoretical  guarantees  in  wide  range  of  settings,  in  several  experiments  involving  non-stationarity,  constraints,  and  multiple-objectives,  our  algorithm  is  able  to  outperform  existing  heuristics  such  as  Thompson  Sampling.The  third  study  develops  a  model-based  reinforcement  learning  approach  for  queueing  network  control  through  differentiable  discrete-event  simulation.  Model-free  RL  methods  have  been  applied  to  queueing  networks,  but  they  are  sample  inefficient  and  suffer  from  slow  convergence.  Despite  non-differentiability  of  the  dynamics  of  queueing  networks,  we  formulate  a  differentiable  approximation  of  the  dynamics  and  use  this  to  obtain  pathwise  policy  gradients  that  are  orders  of  magnitude  less  noisy  than  standard  REINFORCE  gradients.  Using  these  pathwise  policy  gradients,  we  are  able  to  learn  scheduling  policies  that  outperform  or  match  the  policies  trained  with  model-free  RL  with  significantly  less  data.  Unlike  prior  tailored  approaches  to  queueing,  our  methods  can  flexibly  handle  realistic  scenarios,  including  systems  operating  in  non-stationary  environments  and  those  with  non-exponential  interarrival/service  times.The  fourth  chapter  is  a  theoretical  study  of  the  convergence  of  stochastic  gradient  descent  with  Markovian  feedback.  While  there  are  many  works  on  the  convergence  of  policy  gradient  with  REINFORCE-based  gradient  estimation,  in  this  chapter  we  provide  a  convergence  analysis  for  broader  gradient  estimators,  including  differentiable  simulation  or  finite-difference  based  estimators.  The  core  challenge  is  that  the  policy  changes  the  environment  and  thereby  affects  the  data  used  to  update  the  policy.  The  influence  of  previous  decisions  on  the  data  generated  introduces  bias  in  the  gradient  estimate,  which  presents  a  potential  source  of  instability  for  online  learning  not  present  in  the  iid  case.  In  this  paper,  we  introduce  simple  criteria  for  the  adaptively  generated  data  stream  to  guarantee  the  convergence  of  SGD,  which  depend  on  the  mixing  time  of  the  policy-induced  dynamics.  Our  Lyapunov-function  analysis  allows  one  to  translate  existing  stability  analysis  of  stochastic  systems  studied  in  operations  research  into  convergence  rates  for  SGD,  and  we  demonstrate  this  for  queueing  and  inventory  management  problems.  We  also  showcase  how  our  result  can  be  applied  to  study  the  sample  complexity  of  an  actor-critic  policy  gradient  algorithm.
■590    ▼aSchool  code:  0054.
■650  4▼aIndustrial  engineering
■650  4▼aStatistics
■653    ▼aAdaptive  experimentation
■653    ▼aMachine  learning
■653    ▼aQueuing  systems
■653    ▼aReinforcement  learning
■653    ▼aInfinitesimal  perturbation  analysis
■690    ▼a0796
■690    ▼a0546
■690    ▼a0800
■690    ▼a0463
■71020▼aColumbia  University▼bOperations  Research.
■7730  ▼tDissertations  Abstracts  International▼g87-01B.
■790    ▼a0054
■791    ▼aPh.D.
■792    ▼a2025
■793    ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17358479▼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
    TF16052 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.