• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.427-436
• /
• 2000

In this paper, we consider a Bayesian model selection problem of lifetime distributions using fractional Bayes factor with noninformative prior when type II censored data are given. For a given type II censored data, we calculate the posterior probability of exponential, Weibull and lognormal distributions and select the model which gives the highest posterior probability. Our proposed methodology is explained and applied to real data and simulated data.

• The Korean Journal of Applied Statistics
• /
• v.24 no.6
• /
• pp.1213-1224
• /
• 2011

In high dimensionality where the number of variables are excessively larger than observations, it is required to remove the noninformative variables to cluster observations. Most model-based approaches for variable selection have been considered under the assumption of homoscedasticity and their models are mainly estimated by a penalized likelihood method. In this paper, a different approach is proposed to remove the noninformative variables effectively and to cluster based on the modified normal mixture model simultaneously. The validity of the model was provided and an EM algorithm was derived to estimate the parameters. Simulation studies and an experiment using real microarray dataset showed the effectiveness of the proposed method.

• The Korean Journal of Applied Statistics
• /
• v.26 no.1
• /
• pp.141-149
• /
• 2013

We present methods for studying the log-density ratio that enables the selection of the predictors and the form to be included in the logistic regression model. Under bivariate normal distributional assumptions, we investigate the form of the log-density ratio as a function of two predictors. If two covariance matrices are equal, then the crossproduct and quadratic terms are not needed. If the variables are uncorrelated, we do not need the crossproduct terms, but we still need the linear and quadratic terms. We also explore other conditions in which the crossproduct and quadratic terms are not needed in the logistic regression model.

• Journal of Labour Economics
• /
• v.32 no.3
• /
• pp.1-26
• /
• 2009

Using panel data from the Korean Labor and Income Panel Study(KLlPS), we examine the wage differentials between standard and Don-standard workers. To control for unobserved individual heterogeneities, we estimate the fixed effect models. Our results show that the OLS estimates are upwardly biased. We also find that labor unions and firm size are important determinants of the wage differentials.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2006.05a
• /
• pp.1797-1801
• /
• 2006

수문모형은 많은 물리적, 식생적, 기후적, 인위적 요소들의 결과로 기인하는 수문학적 특성을 나타내는 유역의 복잡한 시스템을 현실적으로 표현하는 도구로써 인식되어 왔다. 공간적으로 분포된 수문모형들은 1960년대 처음으로 개발되었으며, 수문학과 수자원관리 분야에서 원격탐사데이터와 지리정보시스템의 그 역할은 급속도록 증가하였다. 비록 원격탐사자료가 수문학분야에 실제 적용된 경우는 매우 적지만, 그 효용성은 크다고 할 수 있다. 수문 모델링과 모니터링분야에서 원격탐사 자료를 이용함에 있어 가장 큰 장점 중의 하나는 시공간적인 정보를 지속적으로 생산할 수 있게 되었다는 점이다. 이와 같은 능력은 성공적인 모형의 분석과 예측, 검증을 위한 작업에 필수적이다. 본 연구는 준 분포형 수문학적 모형인 SLURP 모형을 경안천 유역을 대상으로 적용하였으며, MODIS 위성영상을 이용하여 제작한 엽면적지수(LAI), 정규식생지수(NDVI)를 수문모형의 입력자료로 활용하여 경안 수위표 지점에서 일 유출량 모의를 실시하였다. 또한, 각각의 원격탐사 자료가 모의된 증발산량의 민감도에 어떤 영향을 미치는 가를 분석하였다.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.371-381
• /
• 2000

In this paper, we describes smoothing spline in nonparametric regression and some asymptotic results for estimates of the regression cofficients in the parametric part were biased on semi parametric regression estimator.

• The Korean Journal of Applied Statistics
• /
• v.18 no.2
• /
• pp.381-393
• /
• 2005

As a promising technique for dimension reduction in regression analysis, Sliced Inverse Regression (SIR) and an associated chi-square test for dimensionality were introduced by Li (1991). However, Li's test needs assumption of Normality for predictors and found to be heavily dependent on the number of slices. We will provide a unified asymptotic test for determining the dimensionality of the SIR model which is based on the probabilistic principal component analysis and free of normality assumption on predictors. Illustrative results with simulated and real examples will also be provided.

• Journal of Labour Economics
• /
• v.30 no.2
• /
• pp.1-31
• /
• 2007

The purpose of this paper is to analyse wage differentials between non-regular and regular works. Data from EAPS(Economically Active Population Survey) 2005 show that the monthly wage level of non-regular worker is only 63% of regular worker and thus there exist 37% wage differentials. However, these wage differentials do not control for hours of work, the amount of human capital, job characteristics, and other individual characteristics affecting wages. If these variables are added to the hourly wage regression equation, the wage gap between non-regular and regular workers drastically decreases to 2.2%. Furthermore, decomposition of the wage differentials by Oaxaca method shows that productivity difference between non-regular and regular workers explains up to 91% of the wage gap. This implies that the magnitude of wage discrimination against non-regular workers is at most 0.2% of hourly wage of regular workers. To control for unobserved individual heterogeneities more accurately, we also construct panel data and estimate wage differentials. The results from the panel data approach show that there is no difference in the hourly wages between non-regular and regular workers. In some specifications, the wage rate of non-regular worker is rather higher than that of regular worker. These results are consistent with economic theory. Other things being equal, workers with unstable employment may require higher wages to compensate their unstability. Firms are willing to pay higher wages if they can get more flexibility from non-regular employment. Empirical results in this paper cast doubt on the view that there is wage discrimination against non-regular workers in the labor market. Public policies should be targeted for disadvantaged groups among non-regular workers, not for non-regular workers in general.

• The Korean Journal of Applied Statistics
• /
• v.21 no.5
• /
• pp.825-833
• /
• 2008

Discriminant analysis based on Gaussian mixture models, an useful tool for multi-class classifications, can be extended to semi-supervised learning. We consider a model selection problem for a Gaussian mixture model in semi-supervised learning. More specifically, we adopt Bayesian information criterion to determine the number of subclasses in the mixture model. Through simulations, we illustrate the usefulness of the criterion.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1077-1085
• /
• 2017

In this study, we have found the major factors which affect Korean women's wage analysing the data provided by 2015 Korea Labor Panel Survey (KLIPS). In general, wage data is difficult to analyze because random sampling is infeasible. Heckman sample selection model is the most widely used method for analysing the data with sample selection. Heckman proposed two kinds of selection models: the one is the model with maximum likelihood method and the other is the Heckman two stage model. Heckman two stage model is known to be robust to the normal assumption of bivariate error terms. Recently, Marchenko and Genton (2012) proposed the Heckman selectiont model which generalizes the Heckman two stage model and concluded that Heckman selection-t model is more robust to the error assumptions. Employing the two models, we carried out the analysis of the data and we compared those results.