서브메뉴
검색
Learning From High-Dimensional Measurements
Learning From High-Dimensional Measurements
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20260202103505
- ISBN
- 9798280712003
- DDC
- 310
- 저자명
- Gowri, Gokul.
- 서명/저자
- Learning From High-Dimensional Measurements
- 발행사항
- [Sl] : Harvard University, 2025
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2025
- 형태사항
- 158 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 86-12, Section: B.
- 주기사항
- Advisor: Yin, Peng.
- 학위논문주기
- Thesis (Ph.D.)--Harvard University, 2025.
- 초록/해제
- 요약This thesis considers the problems of distilling and quantifying information in high-dimensional measurements, with a focus on applications in biology. First, we explore the idea that underlying low-dimensional structure in high-dimensional data can be exploited to circumvent the curse of dimensionality in mutual information estimation. We develop a method that we call latent MI (LMI) approximation, which applies a nonparametric MI estimator to low-dimensional representations learned by a simple, theoretically-motivated model architecture. Using several benchmarks, we show that unlike existing techniques, LMI can approximate MI well for variables with 103 dimensions when their dependence structure has low intrinsic dimensionality. Second, we study how measurement noise in data affects the quality of representation learning models. Using an information-theoretic metric of representation quality, we show that model performance scales predictably with molecular under sampling noise in single-cell genomic data. We show that the form of this relationship can be recovered from a simple Gaussian noise model, which provides an intuitive interpretation of the law. Finally, we show that the same scaling relationship emerges in image classification problems, suggesting that noise scaling may be a general phenomenon.
- 일반주제명
- Statistics
- 일반주제명
- Biology
- 일반주제명
- Computer science
- 키워드
- Scaling laws
- 키워드
- Latent MI
- 기타저자
- Harvard University Systems Biology
- 기본자료저록
- Dissertations Abstracts International. 86-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008260126s2025 us c eng d■001000017357390
■00520260202103505
■006m o d
■007cr#unu||||||||
■020 ▼a9798280712003
■035 ▼a(MiAaPQ)AAI32002518
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a310
■1001 ▼aGowri, Gokul.▼0(orcid)0000-0003-0911-6889
■24510▼aLearning From High-Dimensional Measurements
■260 ▼a[Sl]▼bHarvard University▼c2025
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2025
■300 ▼a158 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 86-12, Section: B.
■500 ▼aAdvisor: Yin, Peng.
■5021 ▼aThesis (Ph.D.)--Harvard University, 2025.
■520 ▼aThis thesis considers the problems of distilling and quantifying information in high-dimensional measurements, with a focus on applications in biology. First, we explore the idea that underlying low-dimensional structure in high-dimensional data can be exploited to circumvent the curse of dimensionality in mutual information estimation. We develop a method that we call latent MI (LMI) approximation, which applies a nonparametric MI estimator to low-dimensional representations learned by a simple, theoretically-motivated model architecture. Using several benchmarks, we show that unlike existing techniques, LMI can approximate MI well for variables with 103 dimensions when their dependence structure has low intrinsic dimensionality. Second, we study how measurement noise in data affects the quality of representation learning models. Using an information-theoretic metric of representation quality, we show that model performance scales predictably with molecular under sampling noise in single-cell genomic data. We show that the form of this relationship can be recovered from a simple Gaussian noise model, which provides an intuitive interpretation of the law. Finally, we show that the same scaling relationship emerges in image classification problems, suggesting that noise scaling may be a general phenomenon.
■590 ▼aSchool code: 0084.
■650 4▼aStatistics
■650 4▼aBiology
■650 4▼aComputer science
■653 ▼aHigh-dimensional data
■653 ▼aMutual information
■653 ▼aScaling laws
■653 ▼aSingle-cell genomics
■653 ▼aLatent MI
■690 ▼a0463
■690 ▼a0306
■690 ▼a0984
■690 ▼a0800
■71020▼aHarvard University▼bSystems Biology.
■7730 ▼tDissertations Abstracts International▼g86-12B.
■790 ▼a0084
■791 ▼aPh.D.
■792 ▼a2025
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17357390▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
