• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.5
• /
• pp.645-652
• /
• 2016

The statistical modeling of input random variables is necessary in reliability analysis, reliability-based design optimization, and statistical validation and calibration of analysis models of mechanical systems. In statistical modeling methods, there are the Akaike Information Criterion (AIC), AIC correction (AICc), Bayesian Information Criterion, Maximum Likelihood Estimation (MLE), and Bayesian method. Those methods basically select the best fitted distribution among candidate models by calculating their likelihood function values from a given data set. The number of data or parameters in some methods are considered to identify the distribution types. On the other hand, the engineers in a real field have difficulties in selecting the statistical modeling method to obtain a statistical model of the experimental data because of a lack of knowledge of those methods. In this study, commonly used statistical modeling methods were compared using statistical simulation tests. Their advantages and disadvantages were then analyzed. In the simulation tests, various types of distribution were assumed as populations and the samples were generated randomly from them with different sample sizes. Real engineering data were used to verify each statistical modeling method.

• Journal of the Korea Concrete Institute
• /
• v.22 no.2
• /
• pp.171-177
• /
• 2010

Recently, Performance Based Design (PBD) method has been studied as a next generation structural design method, which enables a designed structure to satisfy the required performance during its service life. One method of deciding whether the required performance has been satisfied is Bayesian method, which has been commonly used in seismic analysis. Generally, it is presented as a conditional probability of exceeding some limit state (i.e., collapse) for a given ground motion. In PBD of concrete mixture design, the same methodology can be applied to assess concrete material performance based on some conditional parameters (i.e. strength, workability, carbonation, etc). In this paper, a detailed explanation of the procedure of drawing satisfaction curve by using Bayesian method based on various material parameters is shown. Also, a discussion of using the developed satisfaction curves for PBD for concrete mixture design is presented.

• The Korean Journal of Applied Statistics
• /
• v.20 no.2
• /
• pp.229-244
• /
• 2007

Among the small area estimation methods, it has been known that hierarchical Bayesian(HB) approach is the most reasonable and effective method. However any model based approaches need good explanatory variables and finding them is the key role in the model based approach. As the lacking of explanatory variables, adopting the spatial terms in the model was introduced. Here in this paper, we evaluate the model based methods with/without spatial terms using the diagnostic methods which were introduced by Brown et al. (2001). And Economic Active Population Survey(2005) is used for data analysis.

• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.9-19
• /
• 1994

Beta-binomial model, which is reparametrized in terms of the mean probability $\mu$ of a positive deagnosis and the $\kappa$ of agreement, is widely used in psychology. When $\mu$ is close to 0, inference about $\kappa$ become difficult because likelihood function becomes constant. We consider Bayesian approach in this case. To apply Bayesian analysis, Gibbs sampler is used to overcome difficulties in integration. Marginal posterior density functions are estimated and Bayesian estimates are derived by using Gibbs sampler and compare the results with the one obtained by using numerical integration.

• Proceedings of the Safety Management and Science Conference
• /
• 2013.04a
• /
• pp.517-531
• /
• 2013

보증 데이터를 통해 제품의 수명 및 형상모수를 추정할 때 최우추정법과 같은 전통적인 통계 분석방법(Classical Statistical Method)을 많이 사용하였다. 그러나 전통적인 통계 분석방법을 통해 수명과 형상모수의 추정 시 표본의 크기가 작거나 불완전한 경우 추정량의 신뢰성이 떨어진다는 단점이 있고 또 누적된 경험과 과거자료를 충분히 이용하지 못하는 단점도 있다. 이러한 문제점을 해결하기 위해 모수의 사전분포를 가정하는 베이지안(Bayesian) 기법의 적용이 필요하다. 하지만 보증 데이터분석에 있어서 베이지안 기법을 이용한 연구는 아직 미흡한 실정이다. 본 연구에서는 수명분포가 와이블 분포를 갖는 보증데이터를 활용하여 모수 추정의 효율성을 비교 분석하고자 한다. 이를 위해 와이블 분포의 모수가 대수정규분포를 따르는 사전분포를 갖는 베이지안 기법과 전통적 통계기법인 생명표법(Actuarial method)을 활용하여 추정량을 도출하고 비교 분석하였다. 이를 통해 충분한 관측 데이터를 확보할 수 없는 경우에 베이지안 기법을 이용한 보증 데이터 분석방법의 성능을 확인하고자 한다.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.15 no.5
• /
• pp.2603-2609
• /
• 2014

Reliability analysis of mechanical systems requires statistical modeling of input random variables such as distribution function types and statistical parameters that affect the performance of the mechanical systems. Some random variables are correlated, but considered as independent variables or wrong assumptions on input random variables have been used. In this paper, joint distributions were modeled using copulas and Bayesian method from limited number of data. To verify the proposed method, statistical simulation tests were carried out for various number of samples and correlation coefficients. As a result, the Bayesian method selected the most probable copula types among candidate copulas even though the candidate copula shapes are similar for low correlations or the number of data is limited. The most probable copulas also yielded similar reliabilities with the true reliability obtained from a true copula, so that it can be concluded that the Bayesian method provides accurate statistical modeling for the reliability analysis.

• The Korean Journal of Applied Statistics
• /
• v.36 no.2
• /
• pp.115-127
• /
• 2023

Multinomial probit model is a popular model for multiclass classification and choice model. Markov chain Monte Carlo (MCMC) method is widely used for estimating multinomial probit model, but its computational cost is high. However, it is well known that variational Bayesian approximation is more computationally efficient than MCMC, because it uses subsets of samples. In this study, we describe multinomial probit model with Gaussian process classification and how to employ variational Bayesian approximation on the model. This study also compares the results of variational Bayesian multinomial probit model to the results of naive Bayes, K-nearest neighbors and support vector machine for the UCI mice protein expression level data.

• The Korean Journal of Applied Statistics
• /
• v.27 no.6
• /
• pp.1029-1038
• /
• 2014

Bayesian experimental design is a useful concept in applied statistics for the design of efficient experiments especially if prior knowledge in the experiment is available. However, a theoretical or numerical approach is not simple to implement. We review the concept of a Bayesian experiment approach for linear and nonlinear statistical models. We investigate relationships between prior knowledge and optimal design to identify Bayesian experimental design process characteristics. A balanced design is important if we do not have prior knowledge; however, prior knowledge is important in design and expert opinions should reflect an efficient analysis. Care should be taken if we set a small sample size with a vague improper prior since both Bayesian design and non-Bayesian design provide incorrect solutions.

• Proceedings of the Korea Concrete Institute Conference
• /
• 2010.05a
• /
• pp.161-162
• /
• 2010

In this study, the impact load on concrete compression member was considered to assess the quantitative damage index. The case study was carried out using the LS-DYNA, on explicit finite element analysis program. The parameters for the case study were impact load angle, slenderness ratio, etc. Using the analysis results, the performance based design method for impact load was developed using Bayesian probabilistic method, which can be applied to reinforced concrete column design for impact loads.

• The Korean Journal of Applied Statistics
• /
• v.22 no.6
• /
• pp.1167-1176
• /
• 2009

Few studies have investigated the quantitative relationship between port ownership structure and port efficiency with mixed results. This paper therefore contributes to the empirical literature by investigating the impact of port privatization on port efficiency using sample data drawn from the world's major ports. Moreover, this study applies the Bayesian approach to estimate the impact of port ownership on port efficiency. We fit Bayesian stochastic frontier model which is introduced by Griffin and Steel (2007) by WinBUGS. World's 25 main ports data are used for analysis. Based on MCMC sampling, we estimate parameters of the model and efficiency index of each ports. Moreover, we add estimates from package Frontier 4.1c in order to compare them with Bayesian results.