서브메뉴
검색
A Unified Lyapunov Framework for Finite-Sample Analysis of Reinforcement Learning Algorithms
A Unified Lyapunov Framework for Finite-Sample Analysis of Reinforcement Learning Algorithms
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202105536
- ISBN
- 9798263397593
- DDC
- 621.3
- 저자명
- Chen, Zaiwei.
- 서명/저자
- A Unified Lyapunov Framework for Finite-Sample Analysis of Reinforcement Learning Algorithms
- 발행사항
- [Sl] : Georgia Institute of Technology, 2022
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2022
- 형태사항
- 297 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 87-06, Section: B.
- 주기사항
- Advisor: Maguluri, Siva Theja;Clarke, John Paul.
- 학위논문주기
- Thesis (Ph.D.)--Georgia Institute of Technology, 2022.
- 초록/해제
- 요약Reinforcement learning is a framework for solving sequential decision-making problems without requiring the environmental model, and is viewed as a promising approach to achieve artificial intelligence. However, there is a huge gap between the empirical successes and the theoretical understanding of reinforcement learning. In this thesis, we make an effort to bridging such gap.More formally, this thesis focuses on designing data-efficient reinforcement learning algorithms and establishing their finite-sample guarantees. Specifically, we aim at answering the following question: suppose we carry out some reinforcement learning algorithm with finite amount of samples (or with finite number of iterations), then what can we say about the performance of the output of the algorithm? The more detailed motivation and the research background are presented in Chapter 1.Part I: Stochastic Approximation. The main body of this thesis is divided into three parts. In the first part of the thesis, we focus on studying the stochastic approximation method. Stochastic approximation is the major workhorse for large-scale optimization and machine learning, and is widely used in reinforcement learning for both algorithm design and algorithm analysis. Therefore, understanding the behavior of SA algorithms is of fundamental interest to the analysis of RL algorithms.In Chapter 2 and Chapter 3, we consider Markovian stochastic approximation under a contractive operator and under a strongly pseudo-monotone operator, and establish their finite-sample guarantees. These two results on stochastic approximation are used in later parts of the thesis to study reinforcement learning algorithms with a tabular representation and with linear function approximation. The main technique we use to analyze those stochastic approximation algorithms is the Lyapunov-drift method. Specifically, we construct novel Lyapunov functions (e.g., generalized Moreau envelope in the case of stochastic approximation under a contraction assumption) to capture the dynamics of the corresponding stochastic approximation algorithms, and control the discretization error and the stochastic error. This enables us to derive the one-step drift inequality, which can be repeatedly used to establish the finite-sample bounds.In Chapter 4, we switch our focus from finite-sample analysis to asymptotic analysis, and characterize the stationary distribution of the centered-scaled iterates of several popular stochastic approximation algorithms. Specifically, we show that for stochastic gradient descent, linear stochastic approximation, and contractive stochastic approximation, the stationary distribution of the centered iterates (after proper scaling) is a Gaussian distribution with mean zero and a covariance matrix being the unique solution of an appropriate Lyapunov equation. For stochastic approximation beyond these three types, we numerically demonstrate that the stationary distribution may not be Gaussian in general. The main technique we used for such asymptotic analysis is also Lyapunov method, where the characteristic function was used as the test function.Part II: Reinforcement Learning with a Tabular Representation. In the second part of this thesis, we focus on reinforcement learning with a tabular representation. The preliminaries of reinforcement learning are presented in Chapter 5.In Chapter 6 and Chapter 7, we consider the TD-learning algorithm for solving the policy evaluation problem, which refers to the problem of estimating the performance of a given policy. Solving the policy evaluation problem is an important intermediate step in the popular actor-critic framework for ultimately finding an optimal policy. More specifically, we consider on-policy TD-learning algorithms such as n-step TD and TD(λ) in Chapter 6. By establishing finite-sample guarantees of n-step TD and TD(λ) as explicit functions of the parameters n and λ, we provide theoretical insight into the open problem about the efficiency of bootstrapping, which is about how to choose the parameters n and λ so that n-step TD and TD(λ) achieve their best performance.In Chapter 7, we study the problem of policy evaluation using off-policy sampling, where the policy used to collect samples and the policy whose value function we aim at estimating is different. We provide finite-sample analysis of a generic off-policy multi-step TD-learning algorithm, which subsumes several popular existing algorithms such as Qπ (λ), Tree-Backup(λ), Retrace(λ), and V -trace as its special cases. In addition, our finite-sample bounds demonstrate a trade-off between the variance (which arises due to the product of the importance sampling ratios) and the bias in the limit point (which arises due to various modifications to the importance sampling ratios). Understanding such bias-variance trade-off is at the heart of off-policy learning.
- 일반주제명
- Computer engineering
- 키워드
- Decision-making
- 기본자료저록
- Dissertations Abstracts International. 87-06B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2022 us c eng d■001000017360497
■00520260202105536
■006m o d
■007cr#unu||||||||
■020 ▼a9798263397593
■035 ▼a(MiAaPQ)AAI32314829
■035 ▼a(MiAaPQ)GeorgiaTech66568
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a621.3
■1001 ▼aChen, Zaiwei.
■24512▼aA Unified Lyapunov Framework for Finite-Sample Analysis of Reinforcement Learning Algorithms
■260 ▼a[Sl]▼bGeorgia Institute of Technology▼c2022
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2022
■300 ▼a297 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 87-06, Section: B.
■500 ▼aAdvisor: Maguluri, Siva Theja;Clarke, John Paul.
■5021 ▼aThesis (Ph.D.)--Georgia Institute of Technology, 2022.
■520 ▼aReinforcement learning is a framework for solving sequential decision-making problems without requiring the environmental model, and is viewed as a promising approach to achieve artificial intelligence. However, there is a huge gap between the empirical successes and the theoretical understanding of reinforcement learning. In this thesis, we make an effort to bridging such gap.More formally, this thesis focuses on designing data-efficient reinforcement learning algorithms and establishing their finite-sample guarantees. Specifically, we aim at answering the following question: suppose we carry out some reinforcement learning algorithm with finite amount of samples (or with finite number of iterations), then what can we say about the performance of the output of the algorithm? The more detailed motivation and the research background are presented in Chapter 1.Part I: Stochastic Approximation. The main body of this thesis is divided into three parts. In the first part of the thesis, we focus on studying the stochastic approximation method. Stochastic approximation is the major workhorse for large-scale optimization and machine learning, and is widely used in reinforcement learning for both algorithm design and algorithm analysis. Therefore, understanding the behavior of SA algorithms is of fundamental interest to the analysis of RL algorithms.In Chapter 2 and Chapter 3, we consider Markovian stochastic approximation under a contractive operator and under a strongly pseudo-monotone operator, and establish their finite-sample guarantees. These two results on stochastic approximation are used in later parts of the thesis to study reinforcement learning algorithms with a tabular representation and with linear function approximation. The main technique we use to analyze those stochastic approximation algorithms is the Lyapunov-drift method. Specifically, we construct novel Lyapunov functions (e.g., generalized Moreau envelope in the case of stochastic approximation under a contraction assumption) to capture the dynamics of the corresponding stochastic approximation algorithms, and control the discretization error and the stochastic error. This enables us to derive the one-step drift inequality, which can be repeatedly used to establish the finite-sample bounds.In Chapter 4, we switch our focus from finite-sample analysis to asymptotic analysis, and characterize the stationary distribution of the centered-scaled iterates of several popular stochastic approximation algorithms. Specifically, we show that for stochastic gradient descent, linear stochastic approximation, and contractive stochastic approximation, the stationary distribution of the centered iterates (after proper scaling) is a Gaussian distribution with mean zero and a covariance matrix being the unique solution of an appropriate Lyapunov equation. For stochastic approximation beyond these three types, we numerically demonstrate that the stationary distribution may not be Gaussian in general. The main technique we used for such asymptotic analysis is also Lyapunov method, where the characteristic function was used as the test function.Part II: Reinforcement Learning with a Tabular Representation. In the second part of this thesis, we focus on reinforcement learning with a tabular representation. The preliminaries of reinforcement learning are presented in Chapter 5.In Chapter 6 and Chapter 7, we consider the TD-learning algorithm for solving the policy evaluation problem, which refers to the problem of estimating the performance of a given policy. Solving the policy evaluation problem is an important intermediate step in the popular actor-critic framework for ultimately finding an optimal policy. More specifically, we consider on-policy TD-learning algorithms such as n-step TD and TD(λ) in Chapter 6. By establishing finite-sample guarantees of n-step TD and TD(λ) as explicit functions of the parameters n and λ, we provide theoretical insight into the open problem about the efficiency of bootstrapping, which is about how to choose the parameters n and λ so that n-step TD and TD(λ) achieve their best performance.In Chapter 7, we study the problem of policy evaluation using off-policy sampling, where the policy used to collect samples and the policy whose value function we aim at estimating is different. We provide finite-sample analysis of a generic off-policy multi-step TD-learning algorithm, which subsumes several popular existing algorithms such as Qπ (λ), Tree-Backup(λ), Retrace(λ), and V -trace as its special cases. In addition, our finite-sample bounds demonstrate a trade-off between the variance (which arises due to the product of the importance sampling ratios) and the bias in the limit point (which arises due to various modifications to the importance sampling ratios). Understanding such bias-variance trade-off is at the heart of off-policy learning.
■590 ▼aSchool code: 0078.
■650 4▼aComputer engineering
■653 ▼aDecision-making
■653 ▼aReinforcement learning
■690 ▼a0464
■71020▼aGeorgia Institute of Technology.
■7730 ▼tDissertations Abstracts International▼g87-06B.
■790 ▼a0078
■791 ▼aPh.D.
■792 ▼a2022
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17360497▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


