• Journal of applied mathematics & informatics
• /
• 제18권1_2호
• /
• pp.247-259
• /
• 2005

Necessary and sufficient conditions, under which there exists (at least) a sequence of vectors of real numbers for which the distribution function (d.f.) of any vector of extreme order statistics converges to a non-degenerate limit, are derived. The interesting thing is that these conditions solely depend on the univariate marginals. Moreover, the limit splits into the product of the limit univariate marginals if all the bivariate marginals of the trivariate d.f., from which the sample is drawn, is of negative quadrant dependent random variables (r.v.'s). Finally, all these results are stated for the multivariate extremes with arbitrary dimensions.

• 응용통계연구
• /
• 제19권3호
• /
• pp.481-504
• /
• 2006

금융위험의 측정 및 관리를 위한 도구로서 분포의 꼬리 부분과 관련한 위험척도로 VaR가 현재 널리 활용되고 있다. 특히 VaR의 정확한 추정을 위해 정규분포를 가정한 기존의 방법보다는 극단치이론을 이용한 방법이 최근 관심을 끌고 있다. 지금까지 극단치이론을 이용한VaR의 추정에 관한 연구는 대부분 단변량의 경우에 대해 이루어졌다. 본 논문에서는 코퓰러를 극단치이론에 결부시켜 다변량 극단치분포를 모형화하여 포트폴리오 위험측정을 다루고 있다. 특히 본 연구에서는 포트폴리오 위험 척도로 VaR와 더불어 ES에 대한 추정 방법도 함께 논의하였다. 포트폴리오 위험측정을 위한 방법으로 본 논문에서 논의한 코퓰러-극단치이론에 의한 접근방법이 기존의 분산-공분산 방법보다 상대적으로 우수한지를 실증자료에 대한 사후검증을 통해 살펴보았다.

• 응용통계연구
• /
• 제23권2호
• /
• pp.219-234
• /
• 2010

시가총액에 따른 인덱스(INDEX) 투자를 했을 경우에, VaR(Value at Risk)을 종합주가지수(KOSPI)로부터 얻은 수익율의 극단 손실값들로부터 추정한다. 이를 위해, 극단값 이론 중 BM(Block Maxima) 모형을 적용하며, 극단 손실값들의 비독립적 발생을 고려하기 위하여, extremal index 역시 추정한다. 모형의 타당성을 알아보기 위해, 실패율방법을 이용한 사후검정 (back-testing) 을 실시한다. 사후검정을 통해, BM 모형을 적용한 VaR의 추정이 적절함을 알 수 있었다. 또한, 일반적으로 많이 사용되는 GARCH 모형을 이용한 VaR의 추정과 비교한다. 이를 통해, 오차가 t-분포를 따른다고 가정하는 경우, GARCH 모형을 이용한 VaR의 추정이 BM 모형을 이용한 경우와 사후 검정결과에 차이가 없음을 확인하였다. 그러나, GARCH 모형을 통한 VaR 추정은 추정시점근방의 극단 손실값들에 민감하게 반응하지만, BM 모형은 그렇지 않았다. 따라서, 현 시점으로부터 단기간동안의 손실위험은 GARCH 모형을 이용한 VaR의 추정값을 사용하는 것이 적절하며, 장기간동안의 손실위험은 BM 모형으로부터 얻은 VaR의 추정값을 사용하는 것이 적절하다.

• Wind and Structures
• /
• 제3권1호
• /
• pp.69-71
• /
• 2000

• Ocean Science Journal
• /
• 제41권4호
• /
• pp.245-254
• /
• 2006

Long term wave climate of both extreme wave and operational wave height is essential for planning and designing coastal structures. Since the field wave data for the waters around Korean peninsula is not enough to provide reliable wave statistics, the wave climate information has been generated by means of long-term wave hindcasting using available meteorological data. Basic data base of hindcasted wave parameters such as significant wave height, peak period and direction has been established continuously for the period of 25 years starting from 1979 and for major 106 typhoons for the past 53 years since 1951 for each grid point of the North East Asia Regional Seas with grid size of 18 km. Wind field reanalyzed by European Center for Midrange Weather Forecasts (ECMWF) was used for the simulation of waves for the extra-tropical storms, while wind field calculated by typhoon wind model with typhoon parameters carefully analyzed using most of the available data was used for the simulation of typhoon waves. Design wave heights for the return period of 10, 20, 30, 50 and 100 years for 16 directions at each grid point have been estimated by means of extreme wave analysis using the wave simulation data. As in conventional methodsi of design criteria estimation, it is assumed that the climate is stationary and the statistics and extreme analysis using the long-term hindcasting data are used in the statistical prediction for the future. The method of extreme statistical analysis in handling the extreme vents like typhoon Maemi in 2003 was evaluated for more stable results of design wave height estimation for the return periods of 30-50 years for the cost effective construction of coastal structures.

• Communications for Statistical Applications and Methods
• /
• 제25권1호
• /
• pp.15-27
• /
• 2018

Parametric method of flood frequency analysis involves fitting of a probability distribution to observed flood data. When record length at a given site is relatively shorter and hard to apply the asymptotic theory, an alternative distribution to the generalized extreme value (GEV) distribution is often used. In this study, we consider the beta-P distribution (BPD) as an alternative to the GEV and other well-known distributions for modeling extreme events of small or moderate samples as well as highly skewed or heavy tailed data. The L-moments ratio diagram shows that special cases of the BPD include the generalized logistic, three-parameter log-normal, and GEV distributions. To estimate the parameters in the distribution, the method of moments, L-moments, and maximum likelihood estimation methods are considered. A Monte-Carlo study is then conducted to compare these three estimation methods. Our result suggests that the L-moments estimator works better than the other estimators for this model of small or moderate samples. Two applications to the annual maximum stream flow of Colorado and the rainfall data from cloud seeding experiments in Southern Florida are reported to show the usefulness of the BPD for modeling hydrologic events. In these examples, BPD turns out to work better than $beta-{\kappa}$, Gumbel, and GEV distributions.

• Structural Engineering and Mechanics
• /
• 제18권5호
• /
• pp.609-626
• /
• 2004

The Land Transport Authority of Singapore has a continuing program of highway bridge upgrading, to refurbish and strengthen bridges to allow for increasing vehicle traffic and increasing axle loads. One subject of this program has been a short span bridge taking a busy highway across a coastal inlet near a major port facility. Experiment-based structural assessments of the bridge were conducted before and after upgrading works including strengthening. Each assessment exercise comprised two separate components; a strain and acceleration monitoring exercise lasting approximately one month, and a full-scale dynamic test carried out in a single day. This paper reports the application of extreme value statistics to estimate bridge live loads using strain measurements.

• Journal of the Korean Data and Information Science Society
• /
• 제25권4호
• /
• pp.903-914
• /
• 2014

The inverse Weibull distribution has been proposed as a model in the analysis of life testing data. Also, inverse Weibull distribution has been recently derived as a suitable model to describe degradation phenomena of mechanical components such as the dynamic components (pistons, crankshaft, etc.) of diesel engines. In this paper, we derive the approximate maximum likelihood estimators of the scale parameter and the shape parameter in the inverse Weibull distribution under multiply type-II censoring. We also develop four modified empirical distribution function (EDF) type tests for the inverse Weibull or extreme value distribution based on multiply type-II censored samples. We also propose modified normalized sample Lorenz curve plot and new test statistic.

• Journal of the Korean Statistical Society
• /
• 제26권3호
• /
• pp.289-297
• /
• 1997

It is well known that the spacings, the differences of two successive order statistics, in a random sample of size n from a distribution function F are independent and exponentially distributed if F is itself the exponential distribution. In this paper we obtain an asymptotically similar result on a fixed number of upper spacings as n .to. .infty. for a general F under the assumption that F is in the domain of attraction of some extreme value distribution. For a heavy or short tailed F, appropriate log transformations of the sample should be proceded to get the result. As a by-product, we also get that each upper spacing diverges in probability to .infty. and converges in probability to 0 as n .to. .infty. for a heavy and short tailed F, respectively, which is fully expected.

• Communications for Statistical Applications and Methods
• /
• 제13권2호
• /
• pp.309-318
• /
• 2006

There are many statistical methods of testing the equality of two population variances. Among them, the well-known F test is very sensitive to the normality assumption. Several other tests that do not assume normality have been proposed, but these tests usually need tables of critical values or software for hypotheses testing. McGrath and Yeh (2005) suggested a quick and compact Count Five test requiring only the calculation of the number of extreme points. Since the Count Five test uses only extreme values, this discards some information from the samples, often resulting in a degradation in power. In this paper, an alternative Count Five test using the trimmed mean is proposed and its properties are discussed for some distributions and normal mixtures.