View : 416 Download: 0

Full metadata record

DC Field Value Language
dc.contributor.advisor김경원-
dc.contributor.author권가윤-
dc.creator권가윤-
dc.date.accessioned2025-02-18T16:32:23Z-
dc.date.available2025-02-18T16:32:23Z-
dc.date.issued2025-
dc.identifier.otherOAK-000000242898-
dc.identifier.urihttps://dcollection.ewha.ac.kr/common/orgView/000000242898en_US
dc.identifier.urihttps://dspace.ewha.ac.kr/handle/2015.oak/271004-
dc.description.abstractSufficient dimension reduction has developed as a powerful tool to extract the core information hidden in high-dimensional data in the past decades. However, as most of the sufficient dimension reduction methods provide linear combinations of the original predictors, interpreting the extracted components becomes challenging. Sparse sufficient dimension reduction was introduced by Li (2007) to facilitate simpler interpretation by producing sparse estimates. In this paper, we introduce sparse directional regression by extending the proposed methods in Li (2007). To demonstrate the competitiveness of our method, we compare the performance of sparse directional regression with that of sparse sliced inverse regression and sliced average variance estimation through numerical experiments. We further apply our method to the large-scale wave energy farm dataset.;충분차원축소는 지난 몇 십년간 고차원 데이터의 중요한 정보를 추출하는 강력한 도구로 개발되었다. 하지만, 대부분의 충분차원축소 방법론들은 기존반응변수의 선형결합으로 나타나기 때문에 추출된 요소를 해석하는 것에 어려움이 따른다. 희소 충분 차원 축소(Sparse sufficient dimensino reduction)는 희소 추정치를 통해 해석을 간단하게 하기 위해 Li (2007)에 의해 도입되었다. 본 논문에서는 Li (2007)에서 제안된 방법을 확장하여 희소 방향 회귀(Sparse directional regression)를 소개한다. 본 방법의 효과를 입증하기 위해 희소 방향 회귀의 성능을 다른 희소 충분 차원 축소 방법론들과 비교한다. 또한, 대규모 파력 발전소 데이터셋에 본 방법을 적용한다.-
dc.description.tableofcontents1 Introduction 1 2 Sparse directional regression 3 2.1 Sliced inverse regression 3 2.2 Sliced average variance estimation 3 2.3 Directional regression 4 2.4 Sparse sufficient dimension reduction 5 2.5 Algorithm 7 3 Simulated examples 7 4 Real data application 14 5 Discussion 16 References 17-
dc.formatapplication/pdf-
dc.format.extent2510128 bytes-
dc.languageeng-
dc.publisher이화여자대학교 대학원-
dc.subjectSufficient dimension reduction, Directional regression, LASSO, High-dimensional data, Multivariate analysis-
dc.subject.ddc500-
dc.titleOn a Sparse and Exhaustive Sufficient Dimension Reduction Method-
dc.typeMaster's Thesis-
dc.creator.othernameKwon, Gayun-
dc.format.pageii, 19 p.-
dc.identifier.thesisdegreeMaster-
dc.identifier.major대학원 통계학과-
dc.date.awarded2025. 2-
Appears in Collections:
일반대학원 > 통계학과 > Theses_Master
Files in This Item:
There are no files associated with this item.
Export
RIS (EndNote)
XLS (Excel)
XML


qrcode

BROWSE