• The Korean Journal of Applied Statistics
• /
• v.27 no.4
• /
• pp.645-653
• /
• 2014

P-value is the probability of observing a current sample and possibly other samples departing equally or more extremely from the null hypothesis toward postulated alternative hypothesis. When p-value is less than a certain level called ${\alpha}$(= 0:05), researchers claim that the alternative hypothesis is supported empirically. Unfortunately, some findings discovered in that way are not reproducible, partly because the p-value itself is a statistic vulnerable to random variation. Boos and Stefanski (2011) suggests calculating the upper limit of p-value in hypothesis testing, using a bootstrap predictive distribution. To determine the sample size of a replication study, this study proposes thought experiments by simulating boosted bootstrap samples of different sizes from given observations. The method is illustrated for the cases of two-group comparison and multiple linear regression. This study also addresses the reproducibility of the points in the given 95% confidence interval. Numerical examples show that the center point is covered by 95% confidence intervals generated from bootstrap resamples. However, end points are covered with a 50% chance. Hence this study draws the graph of the reproducibility rate for each parameter in the confidence interval.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.12 no.20
• /
• pp.25-32
• /
• 1989

열화수신포에 있어서 확률론적 신뢰도설계 및 단순스트레스와 강도의 고장모형은 렌덤 수치들로 구성된 각 분포의 간섭(Interference) 확률에 의해 이루어진다. 현재까지 이러한 분포모형들의 신경도는 분석적, 수치적 또는 모의실험방법론학에 의해서 단일추정치를 계산및 해석하여 왔으나 현실적으로는 스트레스 강도함수 분포의 간섭구간에 있어서 그 렌덤 변수들에 대한 완전한 실제의 분포들을 파악하기란 너무 복잡하고 때로는 불가능하다. 이를 위해서는 신뢰도의 상·하한 결정으로 신뢰도설계를 하는 방법이 고려된다. 본 연구에서는 스트레스와 강도의 렌덤 변수들에 대한 중복부분(Overlapping tails) 내에 있어서 각각의 주어진 구간확률을 제약조건으로 하여 최적신뢰도 한계를 결정 및 설계하는 새로운 기법을 제시한다. 설계신뢰도의 최적상한 및 하한 값은 제공하기 위하여 비선형모형인 Quadratic programming optimization 기법을 이용하여 신뢰도분석관점에 입각한 특수구조모형을 설정하고 이에 대한 수치예를 들어 예증하였다.

• The Korean Journal of Applied Statistics
• /
• v.14 no.2
• /
• pp.401-413
• /
• 2001

이차원 분할자료에 대해서 행과 열의 관계를 저차원상에 시각적으로 표현하는 탐색적대응분석에 대하여 붓스트랩방법의 사용가능성을 살펴보았다. 기존의 탐색적 면만이 강조되어 왔던 대응분석에서 좌표점의 변이와 좌표점간의 거리에 대한 통계적 추론을 붓스트랩방법으로 해결할 수 있음을 보이고 또한 좌표축의 설명력에 대하여 붓스트랩신뢰구간의 포함확률의 일치성을 모의실험을 통해 제시하였다.

• International Journal of Highway Engineering
• /
• v.12 no.1
• /
• pp.61-69
• /
• 2010

In this paper, we present the methodology about an optimal probability distribution selection as well as survival rate estimation with the national highway database from 1999 to 2008. Probability paper methods are adopted to estimate the parameters of each hazard model. The goodness-of-fit test, such as the Anderson-Darling statistics, was performed. As a result, we found that Lognormal distributionan is an appropriate distribution of newly constructed sections as well as overlayed sections. We also ascertained that the results of survival rate for pavement life between the proposed method and observed data are similar. Such a selection methodology and measures based on reliability theory can provide useful information for maintenance plans in pavement management systems as long as additional life data on pavement sections are accumulated.

• Journal of Korean Society of Transportation
• /
• v.24 no.1 s.87
• /
• pp.121-131
• /
• 2006

Considering characteristics of drivers and vehicles with proper and reliable ways in highway design Procedures can ensure high level of highway safety. However, it is almost impossible to take into account all factors of drivers and vehicles influencing on the highway safety because of their uncertain and random nature. To detour the dead-end, the nature are usually assumed as simple homogeneous and deterministic one. Although the restricted assumption makes the system simple, it can produce serious problems due to lack of considering variability in the system. This paper develops a reliability-based method for determining stopping sight distance(SSD) and intersection sight distance (ISD), which are crucial elements in highway alignment design. In the study, Hasofer-Lind method is adopted. which is a well-known first-order second moment reliability method (AFOSM) The results in this study show that if mean, variance, and distribution of a particular driver-vehicle parameter are known, more reliable sight distances can be applied in highway design procedures because we can reflect uncertainties and randomness. Thus, the Probabilistic method could be adopted in designing the sight distance(s) with the desired reliability level.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2009.05a
• /
• pp.1406-1411
• /
• 2009

Bootstrap 기법은 통계학적 추정치의 정확도 또는 불확실성을 평가하기 위한 컴퓨터 기반 리샘플링 기법으로서 플러그인 원칙을 이용하여 요약통계치의 표준오차 및 신뢰구간을 추정하며, Bootstrap 기법 중 BCa 기법은 다른 Bootstrap 기법들에 비해 적합도 기준면에서 훨씬 우수한 결과를 나타내는 것으로 알려져 있다. 본 논문에서는 강우빈도해석에서 확률분포의 매개변수 추정에 대한 불확실성 고려한 확률강우량의 산정 및 불확실성의 영향을 평가하기 위하여 Bootstrap 기법 중 비매개변수적 BCa 기법에 기반한 불확실성을 고려한 강우빈도해석모델 구축 및 적용을 통해 홍수위험평가 및 수자원 계획 등에 있어서 불확실성 표현 및 처리기법을 제시하였다.

• Journal of the Korean Geosynthetics Society
• /
• v.9 no.4
• /
• pp.17-25
• /
• 2010

Evaluation of stability in traditional designing of reinforced soil structures is executed by examination of internal and external stability. Analysis of internal stability is for pull-out and ductile strength. Analysis of external stability is for settlement, overturning and sliding. To minimize inherent uncertainties of soil properties and analytical model, reliability analysis was developed recently. In this study, reliability analysis method considering simultaneous failure probability for various failure mode of internal and external stability is proposed. By applying uni-modal bounds, Stability of system reliability of reinforced soil structures is evaluated by integrating multi failure mode for various analytical model. Because of complex consideration for various failure shapes and modes, it is possible to secure advanced safety by using simultaneous failure probability. And evaluation of reinforced soil structure is executed by representative index, simultaneous failure probability, than previous method.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2004.05b
• /
• pp.863-869
• /
• 2004

댐의 수리${\cdot}$수문학적 월류 확률 추정시에 가장 민감한 불확실성 변량은 댐의 초기수위라 할 수 있으며, 특히 자료의 특성을 충분히 반영하고 댐마루(dam crest)의 높이를 초과하지 않으면서 경계를 갖는 분포형을 추정하는 것은 무엇보다 중요하다. 그러나 기존의 매개변수적 확률분포 추정방법으로 이러한 문제점을 적절히 반영할 수 없으며 통계특성을 반영하지 못하고 이상화시키는 단점이 있다. 이러한 문제점을 보완하기 위해서 비매개변수적 핵밀도함수 방법과 Bootstrap 기법을 적용하여 수위의 신뢰구간을 추정하였다. 연 최대치 자료를 이용한 비매개변수적 핵밀도함수 기법을 이용한 해석결과에서는 댐의 설계빈도를 상회하는 비교적 큰 위험도 나타냈으며 홍수기의 평상수위고 가정하는 Bootstrap Resampling을 적용한 위험도는 5.11E-06의 간을 나타났다. 가장 극심한 기상상태를 가정한 해석 결과인 1.1972E-03은 본 댐은 여수로의 설계빈도가 1,000년 빈도로서 설계당시보다 확률수문량이 크게 증가된 현재 여수로 방류능력 및 안전성 상태로 고려해보면 적당한 위험도 값으로 추정된다.

• Journal of Korea Water Resources Association
• /
• v.50 no.4
• /
• pp.277-288
• /
• 2017

A probabilistic risk analysis of levee system estimates the overall level of flood risk associated with the levee system, according to a series of possible flood scenarios. It requires the uncertainty analysis of all the risk components, including hydrological, hydraulic and geotechnical parts computed by employing MCMC (Markov Chain Monte Carlo), MCS (Monte Carlo Simulation) and FOSM (First-Order Second Moment), presents a joint probability combined each probability. The methodology was applied to a 12.5 km reach from upstream to downstream of the Gangjeong-Goryeong weir, including 6 levee reaches, in Nakdong river. Overtopping risks were estimated by computing flood stage corresponding to 100/200 year high quantile (97.5%) design flood causing levee overflow. Geotechnical risks were evaluated by considering seepage, slope stability, and rapid drawdown along the levee reach without overflow. A probability-based compound risk will contribute to rising effect of safety and economic aspects for levee design, then expect to use the index for riverside structure design in the future.

• Journal of Animal Science and Technology
• /
• v.55 no.3
• /
• pp.179-184
• /
• 2013

Willingness-to-pay (WTP) for pork produced by quality test was determined using a contingent valuation method. Three model studies including a dichotomous-choice and two double dichotomous-choice types were conducted. The respondents in this study appropriately understood the contingent valuation and the suggested price was significant as a characteristic variable. The results imply that there is lower chance to select pork produced by the quality test, as the price difference is greater between conventional and quality-tested pork. WTPs in double and single contingent valuation models were 735 and 547 won/100 g, respectively. WTP was increased with increasing the educational level of respondents. The average WTPs analyzed by convariate were 1,015 won/100 g for double contingent valuation and 580 won/100 g for single contingent valuation. Considering the minimum price of WTP of pork produced by quality test (547.4 won/100 g), the total economic value was estimated to be 5,173,600 million won and per capita customer value was 106,000 won. Therefore, providing an institutional strategy for pork quality test will be beneficial for the consumers.