서브메뉴
검색
Learning Through the Lens of Computation
Learning Through the Lens of Computation
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211151956
- ISBN
- 9798382806686
- DDC
- 004
- 저자명
- Peng, Binghui.
- 서명/저자
- Learning Through the Lens of Computation
- 발행사항
- [Sl] : Columbia University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 245 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
- 주기사항
- Advisor: Papadimitriou, Christos;Chen, Xi.
- 학위논문주기
- Thesis (Ph.D.)--Columbia University, 2024.
- 초록/해제
- 요약One of the major mysteries in science is the remarkable success of machine learning. While its empirical achievements have captivated the field, our theoretical comprehension lags significantly behind. This thesis seeks for advancing our theoretical understanding of learning and intelligence from a computational perspective. By studying fundamental learning, optimization and decision making tasks, we aspire to shed lights on the impact of computation for artificial intelligence.The first part of this thesis concerns the space resource needed for learning. By studying the fundamental role of memory for continual learning, online decision making and convex optimization, we find both continual learning and efficient convex optimization require a lot of memory; while for decision making, exponential savings are possible. More concretely, • We prove there is no memory deduction in continual learning, unless the continual learner takes multiple passes over the sequence of environments; • We prove in order to optimize a convex function in the most iteration efficient way, an algorithm must use a quadratic amount of space;• We show polylogarithmic space is sufficient for making near optimal decisions in an oblivious adversary environment; in a sharp contrast, a quadratic saving is both sufficient and necessary to achieve vanishing regret in an adaptive adversarial environment.The second part of this thesis uses learning as a tool, and resolves a series of long-standing open questions in algorithmic game theory. By giving an exponential faster no-swap-regret learning algorithm, we obtain algorithms that achieve near-optimal computation/communication/iteration complexity for computing an approximate correlated equilibrium in a normal-form game, and we give the first polynomial-time algorithm for computing approximate normal-form correlated equilibria in imperfect information games (including Bayesian and extensive-form games).
- 일반주제명
- Computer science
- 일반주제명
- Computer engineering
- 키워드
- Algorithm
- 키워드
- Complexity
- 키워드
- Game theory
- 키워드
- Machine learning
- 키워드
- Decision making
- 기타저자
- Columbia University Computer Science
- 기본자료저록
- Dissertations Abstracts International. 85-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017162298
■00520250211151956
■006m o d
■007cr#unu||||||||
■020 ▼a9798382806686
■035 ▼a(MiAaPQ)AAI31329144
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a004
■1001 ▼aPeng, Binghui.
■24510▼aLearning Through the Lens of Computation
■260 ▼a[Sl]▼bColumbia University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a245 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-12, Section: B.
■500 ▼aAdvisor: Papadimitriou, Christos;Chen, Xi.
■5021 ▼aThesis (Ph.D.)--Columbia University, 2024.
■520 ▼aOne of the major mysteries in science is the remarkable success of machine learning. While its empirical achievements have captivated the field, our theoretical comprehension lags significantly behind. This thesis seeks for advancing our theoretical understanding of learning and intelligence from a computational perspective. By studying fundamental learning, optimization and decision making tasks, we aspire to shed lights on the impact of computation for artificial intelligence.The first part of this thesis concerns the space resource needed for learning. By studying the fundamental role of memory for continual learning, online decision making and convex optimization, we find both continual learning and efficient convex optimization require a lot of memory; while for decision making, exponential savings are possible. More concretely, • We prove there is no memory deduction in continual learning, unless the continual learner takes multiple passes over the sequence of environments; • We prove in order to optimize a convex function in the most iteration efficient way, an algorithm must use a quadratic amount of space;• We show polylogarithmic space is sufficient for making near optimal decisions in an oblivious adversary environment; in a sharp contrast, a quadratic saving is both sufficient and necessary to achieve vanishing regret in an adaptive adversarial environment.The second part of this thesis uses learning as a tool, and resolves a series of long-standing open questions in algorithmic game theory. By giving an exponential faster no-swap-regret learning algorithm, we obtain algorithms that achieve near-optimal computation/communication/iteration complexity for computing an approximate correlated equilibrium in a normal-form game, and we give the first polynomial-time algorithm for computing approximate normal-form correlated equilibria in imperfect information games (including Bayesian and extensive-form games).
■590 ▼aSchool code: 0054.
■650 4▼aComputer science
■650 4▼aComputer engineering
■653 ▼aAlgorithm
■653 ▼aComplexity
■653 ▼aGame theory
■653 ▼aMachine learning
■653 ▼aDecision making
■690 ▼a0984
■690 ▼a0464
■690 ▼a0800
■71020▼aColumbia University▼bComputer Science.
■7730 ▼tDissertations Abstracts International▼g85-12B.
■790 ▼a0054
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17162298▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.


