• 대한수학회지
• /
• 제53권2호
• /
• pp.315-329
• /
• 2016

In the setting of continuous logic, we study atomless probability spaces and atomless random variable structures. We characterize ${\kappa}$-saturated atomless probability spaces and ${\kappa}$-saturated atomless random variable structures for every infinite cardinal ${\kappa}$. Moreover, ${\kappa}$-saturated and strongly ${\kappa}$-homogeneous atomless probability spaces and ${\kappa}$-saturated and strongly ${\kappa}$-homogeneous atomless random variable structures are characterized for every infinite cardinal ${\kappa}$. For atomless probability spaces, we prove that ${\aleph}_1$-saturation is equivalent to Hoover-Keisler saturation. For atomless random variable structures whose underlying probability spaces are Hoover-Keisler saturated, we prove several equivalent conditions.

• Structural Engineering and Mechanics
• /
• 제86권3호
• /
• pp.349-359
• /
• 2023

This paper presents a technique for determining the optimal number of elements in stochastic finite element analysis based on reliability analysis. Using the change-of-variable perturbation stochastic finite element approach, the probability density function of the dynamic responses of stochastic structures is explicitly determined. This method combines the perturbation stochastic finite element method with the change-of-variable technique into a united model. To further examine the relationships between the random fields, discretization of the random field parameters, such as the variance function and the scale of fluctuation, is also performed. Accordingly, the reliability index is calculated based on the explicit probability density function of responses with Gaussian or non-Gaussian random fields in any number of elements corresponding to the random field discretization. The numerical examples illustrate the effectiveness of the proposed method for a one-dimensional cantilever reinforced concrete column and a two-dimensional steel plate shear wall. The benefit of this method is that the probability density function of responses can be obtained explicitly without the use simulation techniques. Any type of random variable with any statistical distribution can be incorporated into the calculations, regardless of the restrictions imposed by the type of statistical distribution of random variables. Consequently, this method can be utilized as a suitable guideline for the efficient implementation of stochastic finite element analysis of structures, regardless of the statistical distribution of random variables.

• 대한원격탐사학회지
• /
• 제22권3호
• /
• pp.211-219
• /
• 2006

A random forests classifier is applied to multi-sensor data fusion for supervised land-cover classification in order to account for the importance of variable. The random forests approach is a non-parametric ensemble classifier based on CART-like trees. The distinguished feature is that the importance of variable can be estimated by randomly permuting the variable of interest in all the out-of-bag samples for each classifier. Two different multi-sensor data sets for supervised classification were used to illustrate the applicability of random forests: one with optical and polarimetric SAR data and the other with multi-temporal Radarsat-l and ENVISAT ASAR data sets. From the experimental results, the random forests approach could extract important variables or bands for land-cover discrimination and showed reasonably good performance in terms of classification accuracy.

• Communications for Statistical Applications and Methods
• /
• 제16권6호
• /
• pp.1013-1022
• /
• 2009

We in this paper study the almost sure convergence for asymptotically almost negatively associated(AANA) random variable sequences and obtain some new results which extend and improve the result of Jamison et al. (1965) and Marcinkiewicz-Zygumnd strong law types in the form given by Baum and Katz (1965), three-series theorem.

• Journal of the Korean Statistical Society
• /
• 제31권1호
• /
• pp.93-107
• /
• 2002

Random effects generalised linear models are useful for analysing clustered count data in which responses are usually correlated. We propose a Bayesian approach to parameter estimation and variable selection in random effects generalised linear models for count data. A simple Gibbs sampling algorithm for parameter estimation is presented and a simple and efficient variable selection is done by using the Gibbs outputs. An illustrative example is provided.

• 대한기계학회논문집A
• /
• 제33권9호
• /
• pp.870-877
• /
• 2009

This study provides how the Dimension Reduction (DR) method as an efficient technique for reliability analysis can acquire its increased efficiency when it is applied to highly nonlinear problems. In the highly nonlinear engineering systems, 4N+1 (N: number of random variables) sampling is generally recognized to be appropriate. However, there exists uncertainty concerning the standard for judgment of non-linearity of the system as well as possibility of diverse degrees of non-linearity according to each of the random variables. In this regard, this study judged the linearity individually on each random variable after 2N+1 sampling. If high non-linearity appeared, 2 additional sampling was administered on each random variable to apply the DR method. The applications of the proposed sampling to the examples produced the constant results with increased efficiency.

• 대한기계학회:학술대회논문집
• /
• 대한기계학회 2008년도 추계학술대회A
• /
• pp.1162-1168
• /
• 2008

This study provides how the Dimension Reduction (DR) method as an efficient technique for reliability analysis can acquire its increased efficiency when it is applied to highly nonlinear problems. In the highly nonlinear engineering systems, 4N+1 (N: number of random variables) sampling is generally recognized to be appropriate. However, there exists uncertainty concerning the standard for judgment of non-linearity of the system as well as possibility of diverse degrees of non-linearity according to each of the random variables. In this regard, this study judged the linearity individually on each random variable after 2N+1 sampling. If high non-linearity appeared, 2 additional sampling was administered on each random variable to apply the DR method. The applications of the proposed sampling to the examples produced the constant results with increased efficiency.

• 응용통계연구
• /
• 제34권2호
• /
• pp.177-190
• /
• 2021

랜덤포레스트는 여러 의사결정나무 모형들을 융합하여 안정성과 예측력을 높여주기 때문에 종종 사용되는 방법이다. 예측력을 증가시키는 반면 해석의 용이성을 희생하기 때문에 이를 보상하기 위해 변수의 중요도를 제공한다. 변수의 중요도는 랜덤포레스트를 구축할 때 변수가 얼마나 중요한 역할을 하는지를 알려 준다. 그러나 어떤 예측변수가 다른 예측변수들과 상관되어 있을 때 기존 알고리즘의 변수중요도는 왜곡될 수 있다. 상관된 예측변수들의 하향 편향은 예측변수의 중요도를 실제 중요도보다 낮게 측정하게 한다. 우리는 기존 알고리즘을 수정하여 상관 예측변수의 하향 편향을 회복하는 새로운 알고리즘을 제안한다. 제안된 알고리즘의 성능은 모의 자료에 의해 증명되고 실제 자료에 의해 설명된다.

• 대한수학회보
• /
• 제45권3호
• /
• pp.509-522
• /
• 2008

The main aim of this paper is to present some results related to asymptotic behavior of distribution functions of random variables of chi-square type $X^2_N={\Sigma}^N_{i=1}\;X^2_i$ with degrees of freedom N, where N is a positive integer-valued random variable independent on all standard normally distributed random variables $X_i$. Two ways for computing the distribution functions of chi-square type random variables with random degrees of freedom are considered. Moreover, some tables concerning considered distribution functions are demonstrated in Appendix.

• Communications for Statistical Applications and Methods
• /
• 제20권3호
• /
• pp.169-174
• /
• 2013

The ROC curve is drawn with two conditional cumulative distribution functions (or survival functions) of the univariate random variable. In this work, we consider joint cumulative distribution functions of k random variables, and suggest a ROC curve for multivariate random variables. With regard to the values on the line, which passes through two mean vectors of dichotomous states, a joint cumulative distribution function can be regarded as a function of the univariate variable. After this function is modified to satisfy the properties of the cumulative distribution function, a ROC curve might be derived; moreover, some illustrative examples are demonstrated.