• The Korean Journal of Applied Statistics
• /
• v.35 no.1
• /
• pp.131-146
• /
• 2022

Logit models are commonly used to predicting and classifying categorical response variables. Most Bayesian approaches to logit models are implemented based on the Metropolis-Hastings algorithm. However, the algorithm has disadvantages of slow convergence and difficulty in ensuring adequacy for the proposal distribution. Therefore, we use auxiliary mixture sampler proposed by Frühwirth-Schnatter and Frühwirth (2007) to estimate logit models. This method introduces two sequences of auxiliary latent variables to make logit models satisfy normality and linearity. As a result, the method leads that logit model can be easily implemented by Gibbs sampling. We applied the proposed method to diabetes data from the Community Health Survey (2020) of the Korea Disease Control and Prevention Agency and compared performance with Metropolis-Hastings algorithm. In addition, we showed that the logit model using auxiliary mixture sampling has a great classification performance comparable to that of the machine learning models.

• Structural Engineering and Mechanics
• /
• v.25 no.3
• /
• pp.347-363
• /
• 2007

A novel methodology, referred to as Auxiliary Domain Method (ADM), allowing for a very efficient solution of nonlinear reliability problems is presented. The target nonlinear failure domain is first populated by samples generated with the help of a Markov Chain. Based on these samples an auxiliary failure domain (AFD), corresponding to an auxiliary reliability problem, is introduced. The criteria for selecting the AFD are discussed. The emphasis in this paper is on the selection of the auxiliary linear failure domain in the case where the original nonlinear reliability problem involves multiple objectives rather than a single objective. Each reliability objective is assumed to correspond to a particular response quantity not exceeding a corresponding threshold. Once the AFD has been specified the method proceeds with a modified subset simulation procedure where the first step involves the direct simulation of samples in the AFD, rather than standard Monte Carlo simulation as required in standard subset simulation. While the method is applicable to general nonlinear reliability problems herein the focus is on the calculation of the probability of failure of nonlinear dynamical systems subjected to Gaussian random excitations. The method is demonstrated through such a numerical example involving two reliability objectives and a very large number of random variables. It is found that ADM is very efficient and offers drastic improvements over standard subset simulation, especially when one deals with low probability failure events.

• Proceedings of the Korean Association for Survey Research Conference
• /
• 2003.06a
• /
• pp.19-28
• /
• 2003

Weighting adjustment is a method of improving the efficiency of the estimator by incorporating auxiliary variables at the estimation stage. One commonly used method of weighting adjustment is the poststratification, which is a special case of regression estimation but is relatively feasible in terms of actual implementation. If too many auxiliary variables are used in the poststratification, the bias of the resulting point estimator is no longer negligible and the final weights may have extreme weights. In this study, we propose a method of weight ing adjustment that compromises the efficiency and the bias of the point estimator. A limited simulation study is also presented.

• Communications of the Korean Mathematical Society
• /
• v.37 no.1
• /
• pp.177-193
• /
• 2022

The main aim of this paper is to introduce a new generalization of the Wright series in two variables, which is expressed in terms of Hermite polynomials. The properties of the freshly defined function involving its auxiliary functions and the integral representations are established. Furthermore, a Gauss-Hermite quadrature and Gaussian quadrature formulas have been established to evaluate some integral representations of our main results and compare them with our theoretical evaluations using graphical simulations.

• Communications for Statistical Applications and Methods
• /
• v.24 no.3
• /
• pp.255-269
• /
• 2017

This article explores the analysis of longitudinal surveys in which same units are investigated on several occasions. Multivariate exponential ratio type estimator has been proposed for the estimation of the finite population median at the current occasion in two occasion longitudinal surveys. Information on several additional auxiliary variables, which are stable over time and readily available on both the occasions, has been utilized. Properties of the proposed multivariate estimator, including the optimum replacement strategy, are presented. The proposed multivariate estimator is compared with the sample median estimator when there is no matching from a previous occasion and with the exponential ratio type estimator in successive sampling when information is available on only one additional auxiliary variable. The merits of the proposed estimator are justified by empirical interpretations and validated by a simulation study with the help of some natural populations.

• Journal of the Korean Statistical Society
• /
• v.35 no.4
• /
• pp.377-395
• /
• 2006

In this paper we have suggested a family of chain estimators of the population mean $\bar{Y}$ of a study variate y using two auxiliary variates in two phase (double) sampling assuming that the coefficient of variation of the second auxiliary variable is known. It is well known that chain estimators are traditionally formulated when the population mean $\bar{X}_1$ of one of the two auxiliary variables, say $x_1$, is not known but the population mean $\bar{X}_2$ of the other auxiliary variate $x_2$ is available and $x_1$ has higher degree of positive correlation with the study variate y than $x_2$ has with y, $x_2$ being closely related to $x_1$. Here the classes are constructed when the population mean $\bar{X}_1\;of\;X_1$ is not known and the coefficient of variation $C_{x2}\;of\;X_2$ is known instead of population mean $\bar{X}_2$. Asymptotic expressions for the bias and mean square error (MSE) of the suggested family have been obtained. An asymptotic optimum estimator (AOE) is also identified with its MSE formula. The optimum sample sizes of the preliminary and final samples have been derived under a linear cost function. An empirical study has been carried out to show the superiority of the constructed estimator over others.

• Survey Research
• /
• v.10 no.1
• /
• pp.155-168
• /
• 2009

As an efficient sampling design, stratified random sampling is often used when auxiliary information is available at the designing stage. Although one - per - stratum design is an efficient design that can be used when many auxiliary variables are available, it does not provide any unbiased variance estimator. With a two - per - stratum sample in which two elements are selected from each stratum, it is possible to obtain an unbiased variance estimator. However the loss of efficiency could be significant if any important stratification variable is missed. In this study, we investigated a sampling design that uses the all given auxiliary information and also permits an unbiased variance estimator suggested by Park and Fuller(2008). Through a simulation study, we compared several stratified random sampling and systematic sampling design. We also applied the proposed stratified sampling designs to 2007 youth panel data.

• Proceedings of the Korean Society of Coastal and Ocean Engineers Conference
• /
• 2004.08a
• /
• pp.310-319
• /
• 2004

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.12 no.3
• /
• pp.133-141
• /
• 2007

A finite element model is developed for the extended Boussinesq equations that is capable of simulating the dynamics of long and short waves. Galerkin weighted residual method and the introduction of auxiliary variables for 3rd spatial derivative terms in the governing equations are used for the model development. The Adams-Bashforth-Moulton Predictor Corrector scheme is used as a time integration scheme for the extended Boussinesq finite element model so that the truncation error would not produce any non-physical dispersion or dissipation. This developed model is applied to the problems of solitary wave propagation. Predicted results is compared to available analytical solutions and laboratory measurements. A good agreement is observed.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.543-553
• /
• 2016

The validity of statistical inference depends on proper randomization methods. However, even with proper randomization, we can have imbalanced with respect to important characteristics. In this paper, we introduce a method based on ranked auxiliary variables for treatment allocation in crossover designs using Latin squares models. We evaluate the improvement of the efficiency in treatment comparisons using the proposed method. Our simulation study reveals that our proposed method provides a more powerful test compared to simple randomization with the same sample size. The proposed method is illustrated by conducting an experiment to compare two different concentrations of titanium dioxide nanofiber (TDNF) on rats for the purpose of comparing weight gain.