View : 685 Download: 0

Full metadata record

DC Field Value Language
dc.contributor.author박수진-
dc.creator박수진-
dc.date.accessioned2016-08-25T02:08:10Z-
dc.date.available2016-08-25T02:08:10Z-
dc.date.issued1997-
dc.identifier.otherOAK-000000026132-
dc.identifier.urihttps://dspace.ewha.ac.kr/handle/2015.oak/174576-
dc.identifier.urihttp://dcollection.ewha.ac.kr/jsp/common/DcLoOrgPer.jsp?sItemId=000000026132-
dc.description.abstract관측자료가 우측 절단자료(right-censored data)였을 경우에 대한 회귀계수의 M-estimators를 계산하는 algorithms을 유도하여 전개하였다. M-estimators를 구함에 있어서는, concomitant scale 추정 방법을 고려하여 결합하는 새로운 통계적 방법을 제시하였다. 모의실험에서 Huber의 M-estimators와 Buckley-James의 추정치를 비교했으며, 그 결과 Huber의 추정치가 보다 효율적이라는 결론을 얻었다.;Computational algorithms to calculate M-estimators of regression parameters from right-censored data are developed herein. In the case of M-estimators, a new statistical method is also introduced to incorporate concomitant scale estimation in the presence of right censoring on the observed responses. Furthermore, we illustrate this by simulations.-
dc.description.tableofcontentsCONTENTS ABSTRACT CHAPTER 1 INTRODUCTION = 1 CHAPTER 2 INITIALIZING WITH SIMPLE PRELIMINARY ESTIMATOR = 4 CHAPTER 3 M-ESTIMATORS AND CONCOMITANT SCALE ESTIMATION = 7 3.1 INCORPORATING SCALE AND DAMPENING INSTABILITY DUE TO SMALL RISK SET SIZES = 8 3.2 COMPUTATION OF M-ESTIMATORS = 11 3.3 SIMULATIONS = 15 CHAPTER 4 CONCLUSION = 19 REFERENCE = 20 논문초록 = 22 감사의 글 = 23-
dc.formatapplication/pdf-
dc.format.extent769304 bytes-
dc.languageeng-
dc.publisher이화여자대학교 대학원-
dc.subjectROBUST REGRESSION-
dc.subject우측절단자료-
dc.subject회귀계수-
dc.subjectRIGHT-CENSORED DATA-
dc.subjectCONCOMITANT SCALE ESTIMATION-
dc.titleROBUST REGRESSION WITH CONCOMITANT SCALE ESTIMATION FOR RIGHT-CENSORED DATA-
dc.typeMaster's Thesis-
dc.format.page24 p.-
dc.identifier.thesisdegreeMaster-
dc.identifier.major대학원 통계학과-
dc.date.awarded1997. 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