• Journal of the Korean Data and Information Science Society
• /
• 제27권5호
• /
• pp.1119-1131
• /
• 2016

Neyman-Scott 구형펄스모형 (Neyman-Scott rectangular pulses model; NSRPM)은 강우의 발생, 강우세포의 강우강도 그리고 지속시간으로 표현되는 점과정에 기초한 강우생성 모형으로, 기존의 구형펄스모형 (rectangular pulse model)과 비교해서 강우사상의 군집특성을 잘 반영하기 때문에 여러 연구에서 많이 사용되는 모형이다. 하지만 NSRPM의 매개변수를 추정하는데 있어서 모멘트를 이용한 여러가지 최적화 기법들은 그 계산이 복잡하고 또한 목적함수의 구성에 따라 추정값의 변동도 크게 나타난다. 이를 보완하기 위해서, 최근 누적강수량에 대한 근사적인 우도함수 (approximated likelihood function)와 이를 통해 NSRPM의 매개변수를 추정하는 방법이 소개되었다. 본 논문에선 이 근사적 우도함수를 바탕으로 계층적 베이지안 모형을 이용하여 NSRPM에 공간구조를 표현하고 이를 통해 강우생성 모형의 공간적 특성을 알아보고자 한다.

• Safety and Health at Work
• /
• 제9권1호
• /
• pp.42-52
• /
• 2018

Background: Maintenance operations on-board ships are highly demanding. Maintenance operations are intensive activities requiring high man-machine interactions in challenging and evolving conditions. The evolving conditions are weather conditions, workplace temperature, ship motion, noise and vibration, and workload and stress. For example, extreme weather condition affects seafarers' performance, increasing the chances of error, and, consequently, can cause injuries or fatalities to personnel. An effective human error probability model is required to better manage maintenance on-board ships. The developed model would assist in developing and maintaining effective risk management protocols. Thus, the objective of this study is to develop a human error probability model considering various internal and external factors affecting seafarers' performance. Methods: The human error probability model is developed using probability theory applied to Bayesian network. The model is tested using the data received through the developed questionnaire survey of >200 experienced seafarers with >5 years of experience. The model developed in this study is used to find out the reliability of human performance on particular maintenance activities. Results: The developed methodology is tested on the maintenance of marine engine's cooling water pump for engine department and anchor windlass for deck department. In the considered case studies, human error probabilities are estimated in various scenarios and the results are compared between the scenarios and the different seafarer categories. The results of the case studies for both departments are also compared. Conclusion: The developed model is effective in assessing human error probabilities. These probabilities would get dynamically updated as and when new information is available on changes in either internal (i.e., training, experience, and fatigue) or external (i.e., environmental and operational conditions such as weather conditions, workplace temperature, ship motion, noise and vibration, and workload and stress) factors.

• 한국지구물리탐사학회:학술대회논문집
• /
• 한국지구물리탐사학회 2003년도 Proceedings of the international symposium on the fusion technology
• /
• pp.390-397
• /
• 2003

A three-dimensional (3D) inversion technique has been developed for interpretation of magnetotelluric (MT) data. The inversion method is based on the linearized least-squares (Gauss-Newton) method with smoothness regularization. In addition to the underground 3D resistivity distribution, static shifts are also treated as unknown parameters in the inversion. The forward modeling is by the staggered-grid finite difference method. A Bayesian criterion ABle is applied to search the optimum trade-off among the minimization of the data misfit, model roughness and static shifts. The method has been applied to several MT datasets obtained at geothermal fields in Japan and other Asian countries. In this paper, two examples will be discussed: one is the data at the Ogiri geothermal area, southwestern Japan, and the other is at the Pohang low-enthalpy geothermal field, southeastern Korea. The inversion of the Ogiri data has been performed stably, resulting in a good fitting between the observed and computed apparent resistivities and phases. The recovered 3D resistivity structure is generally similar to the two-dimensional (2D) inversion models, although the deeper portion of the 3D model seems to be more realistic than that of the 2D model. The 3D model is also in a good agreement with the geological model of the geothermal reservoirs. 3D interpretation of the Pohang MT data is still preliminary. Although the fitting to the observed data is very good, the preliminary 3D model is not reliable enough because the station coverage is not sufficient for a 3D inversion.

• 정보처리학회논문지D
• /
• 제11D권6호
• /
• pp.1269-1276
• /
• 2004

본 논문에서는 기존의 소프트웨어 신뢰성 모형인 Goel-Okumoto 모형과 Yamada-Ohba-Osaki 모형을 재조명하고 또, 랄리 분포를 이용한 랄리 모형을 적용하여 모수 추정방법을 연구하였다. 본 연구에서는 기존의 최우추정법과 잠재변수를 도입하여 깁스 샘플링(Gibbs sampling)을 이용한 베이지안 모수추정 방법을 비교하고 그 특징을 분석하고자 한다. 또, 효율적 모형을 위한 모형선택으로서 잔차제곱합(Sum of the squared errors ; SSE)과 Braun 통계량을 적용하여 모형들에 대한 효율성 입증방법을 설명하였다. 그리고 수치적인 예로서 실제 자료를 이용한 수치 견과를 나열하였다. 이 접근방법을 기초로 하여 수명분포가 중첩(Superposition) 및 혼합(Mixture)인 경우에 대한 접근방법이 연구되었으면 한다.

• 한국지진공학회논문집
• /
• 제25권3호
• /
• pp.145-152
• /
• 2021

KAERI has planned to carry out a series of dynamic tests using a shaking table and time-history analyses for a channel-type concrete shear wall to investigate its seismic performance because of the recently frequent occurrence of earthquakes in the south-eastern parts of Korea. The overall size of a test specimen is b×l×h =2500 mm×3500 mm×4500 mm, and it consists of three stories having slabs and walls with thicknesses of 140 mm and 150 mm, respectively. The system identification, FE model updating, and time-history analysis results for a test shear wall are presented herein. By applying the advanced system identification, so-called pLSCF, the improved modal parameters are extracted in the lower modes. Using three FE in-house packages, such as FEMtools, Ruaumoko, and VecTor4, the eigenanalyses are made for an initial FE model, resulting in consistency in eigenvalues. However, they exhibit relatively stiffer behavior, as much as 30 to 50% compared with those extracted from the test in the 1st and 2nd modes. The FE model updating is carried out to consider the 6-dofs spring stiffnesses at the wall base as major parameters by adopting a Bayesian type automatic updating algorithm to minimize the residuals in modal parameters. The updating results indicate that the highest sensitivity is apparent in the vertical translational springs at few locations ranging from 300 to 500% in variation. However, their changes seem to have no physical meaning because of the numerical values. Finally, using the updated FE model, the time-history responses are predicted by Ruaumoko at each floor where accelerometers are located. The accelerograms between test and analysis show an acceptable match in terms of maximum and minimum values. However, the magnitudes and patterns of floor response spectra seem somewhat different because of the slightly different input accelerograms and damping ratios involved.

• 한국지능시스템학회:학술대회논문집
• /
• 한국지능시스템학회 2007년도 추계학술대회 학술발표 논문집
• /
• pp.63-66
• /
• 2007

기존의 홈네트워크 시스템에서 사용자의 단순한 명령을 통해 서비스를 제공하는 기술은 이미 구현되어 있다. 그렇지만 가정이라는 환경은 이렇게 단순한 환경이기보다, 다수의 가족 구성원으로 이루어져 있으며 그에 따른 다양한 명령과 상황이 존재하고 있다. 이러한 다변화된 특성에 맞추어 사용자의 단순 명령보다 한 단계 높은 수준으로 사용자의 욕구를 능동적으로 추론해 낼 수 있는 모델의 제안이 필요하다. 본 논문에서 베이지안 네트워크를 활용하여 사용자의 주변 환경 정보로 규정된 Context를 인식하고 인식된 결과를 통해 사용자가 요구하는 적합한 서비스(Context-based Service)를 추론해 낼 수 있는 모델을 제시하고자 한다.

• Journal of the Korean Data and Information Science Society
• /
• 제4권
• /
• pp.65-74
• /
• 1993

Suppose that we have two populations(or systems), say ${\Pi}_{1}\;and\;{\Pi}_{2}$, to be tested. A random sample of size n from each population is taken and the test for each system will be terminated when the first r failures among n random samples are observed. This kind of test is caned the type-II censored (or item-censored) testing without replacement. Under this scheme we consider the problem of estimating the unknown parameters of interests and the reliability for a given time t for each population.

• 한국데이터정보과학회:학술대회논문집
• /
• 한국데이터정보과학회 2003년도 추계학술대회
• /
• pp.39-47
• /
• 2003

The Bayes factors with improper noninformative priors are defined only up to arbitrary constants. So, it is known that Bayes factors are not well defined due to this arbitrariness in Bayesian hypothesis testing and model selections. The intrinsic Bayes factor by Berger and Pericchi (1996) and the fractional Bayes factor by O'Hagan (1995) have been used to overcome this problems. This paper suggests intrinsic priors for testing the equality of two lognormal means, whose Bayes factors are asymptotically equivalent to the corresponding fractional Bayes factors. Using proposed intrinsic priors, we demonstrate our results with a simulated dataset.

• Journal of the Korean Data and Information Science Society
• /
• 제22권1호
• /
• pp.125-136
• /
• 2011

Ground-level ozone, an air pollutant that is monitored by the Environmental Protection Agency (EPA), damages human health by irritating the respiratory system, reducing lung function, damaging lung cells, and aggravating asthma and other chronic conditions. In March 2008, the EPA strengthened ozone standards by lowering acceptable limits from 84 parts per billion to 75 parts per billion. Here epidemiologic data is used to study the effects of ozone regulation on human health and assessed how various regulatory standards for ozone may affect nonaccidental mortality, including respiratory-related deaths during ozone season. The assessment uses statistical methods based on hierarchical Bayesian models to predict the potential effects of the different regulatory standards. It also analyzes the variability of the results and ho they are impacted by different modeling assumptions. We focused on the technical an statistical approach to assessing relationship between new ozone regulations and mortality while other researches have detailed the relationship between ozone and human mortality. We shows a statistical correlation between ozone regulations and mortality, with lower limits of acceptable ozone linked to a decrease in deaths, and projects that mortality is expected to decrease by reducing ozone regulatory standards.

• 한국생물정보학회:학술대회논문집
• /
• 한국생물정보시스템생물학회 2005년도 BIOINFO 2005
• /
• pp.357-360
• /
• 2005

In this paper we consider the well-known semiparametric proportional hazards (PH) models for survival analysis. These models are usually used with few covariates and many observations (subjects). But, for a typical setting of gene expression data from DNA microarray, we need to consider the case where the number of covariates p exceeds the number of samples n. For a given vector of response values which are times to event (death or censored times) and p gene expressions (covariates), we address the issue of how to reduce the dimension by selecting the significant genes. This approach enable us to estimate the survival curve when n < < p. In our approach, rather than fixing the number of selected genes, we will assign a prior distribution to this number. The approach creates additional flexibility by allowing the imposition of constraints, such as bounding the dimension via a prior, which in effect works as a penalty. To implement our methodology, we use a Markov Chain Monte Carlo (MCMC) method. We demonstrate the use of the methodology to diffuse large B-cell lymphoma (DLBCL) complementary DNA(cDNA) data.