View : 568 Download: 0

A Monte Carlo study on a new test for unit roots in a seasonal autoregressive model

Title
A Monte Carlo study on a new test for unit roots in a seasonal autoregressive model
Authors
오유진
Issue Date
1997
Department/Major
대학원 통계학과
Keywords
Monte Carlo시계열 자료OLSEmedian-unbiased
Publisher
The Graduate school of Ewha Women's University
Degree
Master
Abstract
시계열 자료에서 단위근 검정시 OLSE(ordinary least square estimator)에 근거한 통계량을 많이 사용해왔다. 본 논문에서는 OLSE에 근거한 통계량의 단점을 보완하여 Cauchy-type estimator라 명명한 추정량을 제시하였다. 제시된 추정량은 median-unbiased 되어있고, 그 통계량은 극한 분포가 정규분포라는 장점이 있다. 그리고 Monte-Carlo simulation을 보면 부분적으로 OLSE에 근거한 테스트보다 부분적으로 더 powerful 하다.;For a seasonal autoregressive process, we propose a new estimator whose pivotal statistic has a standard normal limiting distribution for all range of autoregressive parameter . The proposed estimator is approximately median-unbiased. For seasonal time series, the new estimator gives us unit root tests which has the limiting normal distribution. A Monte-Carlo simulation shows that the proposed tests for unit root are locally more powerful than the tests based on the ordinary least squares estimator.
Fulltext
Show the fulltext
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