서브메뉴
검색
Problems in Variable Selection: False Discovery Rate Control and Variational Inference
Problems in Variable Selection: False Discovery Rate Control and Variational Inference
상세정보
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211151422
- ISBN
- 9798382777672
- DDC
- 310
- 저자명
- Lin, Buyu.
- 서명/저자
- Problems in Variable Selection: False Discovery Rate Control and Variational Inference
- 발행사항
- [Sl] : Harvard University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 316 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
- 주기사항
- Advisor: Liu, Jun.
- 학위논문주기
- Thesis (Ph.D.)--Harvard University, 2024.
- 초록/해제
- 요약Variable selection plays a key role in modern high-dimensional statistics. This dissertation provides a comprehensive survey of theory and methods developed by the author and collaborators, with a specific focus on two areas: false discovery rate (FDR) control and Bayesian variable selection. In the domain of FDR control, we introduce a data-splitting method to asymptotically control the FDR while maintaining a high power. Furthermore, a Multiple Data Splitting (MDS) method is proposed to stabilize the selection result and boost the power. In Chapter 1, we apply both DS and MDS to the generalized linear models, which appear to be more robust in finite-sample cases compared to existing methods. Chapter 2 provides some following discussions regarding the proposed method and Chapter 3 compares the power of the proposed method with two existing methods: the model-X knockoff and Gaussian mirror. In terms of the Bayesian variable selection, the posterior is typically high-dimensional and analytically intractable. Exact inference methods based on sampling, such as Markov Chain Monte Carlo (MCMC), can encounter challenges related to mixing. Variational inference has emerged as an attractive alternative for approximating the posterior distribution. By recasting the sampling problem as an optimization problem, variational inference can significantly reduce computational time. In chapter 4, we apply the variational inference to group variable selection with spike-and-slab prior and propose an efficient parameter-expanded coordinate-ascent algorithm to obtain the optimal variational Bayes approximation. The proposed method has demonstrated good performance in both simulations and a real data example.
- 일반주제명
- Statistics
- 일반주제명
- Endocrinology
- 일반주제명
- Biostatistics
- 키워드
- FDR control
- 기타저자
- Harvard University Statistics
- 기본자료저록
- Dissertations Abstracts International. 85-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017161622
■00520250211151422
■006m o d
■007cr#unu||||||||
■020 ▼a9798382777672
■035 ▼a(MiAaPQ)AAI31294278
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a310
■1001 ▼aLin, Buyu.▼0(orcid)0009-0000-2569-3753
■24510▼aProblems in Variable Selection: False Discovery Rate Control and Variational Inference
■260 ▼a[Sl]▼bHarvard University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a316 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-12, Section: B.
■500 ▼aAdvisor: Liu, Jun.
■5021 ▼aThesis (Ph.D.)--Harvard University, 2024.
■520 ▼aVariable selection plays a key role in modern high-dimensional statistics. This dissertation provides a comprehensive survey of theory and methods developed by the author and collaborators, with a specific focus on two areas: false discovery rate (FDR) control and Bayesian variable selection. In the domain of FDR control, we introduce a data-splitting method to asymptotically control the FDR while maintaining a high power. Furthermore, a Multiple Data Splitting (MDS) method is proposed to stabilize the selection result and boost the power. In Chapter 1, we apply both DS and MDS to the generalized linear models, which appear to be more robust in finite-sample cases compared to existing methods. Chapter 2 provides some following discussions regarding the proposed method and Chapter 3 compares the power of the proposed method with two existing methods: the model-X knockoff and Gaussian mirror. In terms of the Bayesian variable selection, the posterior is typically high-dimensional and analytically intractable. Exact inference methods based on sampling, such as Markov Chain Monte Carlo (MCMC), can encounter challenges related to mixing. Variational inference has emerged as an attractive alternative for approximating the posterior distribution. By recasting the sampling problem as an optimization problem, variational inference can significantly reduce computational time. In chapter 4, we apply the variational inference to group variable selection with spike-and-slab prior and propose an efficient parameter-expanded coordinate-ascent algorithm to obtain the optimal variational Bayes approximation. The proposed method has demonstrated good performance in both simulations and a real data example.
■590 ▼aSchool code: 0084.
■650 4▼aStatistics
■650 4▼aEndocrinology
■650 4▼aBiostatistics
■653 ▼aMultiple Data Splitting
■653 ▼aFalse discovery rate
■653 ▼aMarkov Chain Monte Carlo
■653 ▼aVariational inference
■653 ▼aFDR control
■690 ▼a0463
■690 ▼a0409
■690 ▼a0308
■71020▼aHarvard University▼bStatistics.
■7730 ▼tDissertations Abstracts International▼g85-12B.
■790 ▼a0084
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17161622▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
