• Journal of the Korean Data and Information Science Society
• /
• v.23 no.4
• /
• pp.739-748
• /
• 2012

In this paper we propose a mixed effects least squares support vector machine (LS-SVM) for the censored data which are observed from different groups. We use weights by which the randomly right censoring is taken into account in the nonlinear regression. The weights are formed with Kaplan-Meier estimates of censoring distribution. In the proposed model a random effects term representing inter-group variation is included. Furthermore generalized cross validation function is proposed for the selection of the optimal values of hyper-parameters. Experimental results are then presented which indicate the performance of the proposed LS-SVM by comparing with a standard LS-SVM for the censored data.

• Journal of Preventive Medicine and Public Health
• /
• v.50 no.6
• /
• pp.377-385
• /
• 2017

Objectives: Although mercury (Hg) exposure is known to be neurotoxic in humans, its effects on liver function have been less often reported. The aim of this study was to investigate whether total Hg exposure in Korean adults was associated with elevated serum levels of the liver enzymes aspartate aminotransferase (AST), alanine transaminase (ALT), and gamma-glutamyltransferase (GGT). Methods: We repeatedly examined the levels of total Hg and liver enzymes in the blood of 508 adults during 2010-2011 and 2014-2015. Cross-sectional associations between levels of blood Hg and liver enzymes were analyzed using a generalized linear model, and nonlinear relationships were analyzed using a generalized additive mixed model. Generalized estimating equations were applied to examine longitudinal associations, considering the correlations of individuals measured repeatedly. Results: GGT increased by 11.0% (95% confidence interval [CI], 4.5 to 18.0%) in women and 8.1% (95% CI, -0.5 to 17.4%) in men per doubling of Hg levels, but AST and ALT were not significantly associated with Hg in either men or women. In women who drank more than 2 or 3 times per week, AST, ALT, and GGT levels increased by 10.6% (95% CI, 4.2 to 17.5%), 7.7% (95% CI, 1.1 to 14.7%), and 37.5% (95% CI,15.2 to 64.3%) per doubling of Hg levels, respectively, showing an interaction between blood Hg levels and drinking. Conclusions: Hg exposure was associated with an elevated serum concentration of GGT. Especially in women who were frequent drinkers, AST, ALT, and GGT showed a significant increase, with a significant synergistic effect of Hg and alcohol consumption.

• Proceedings of the Korea Water Resources Association Conference
• /
• 1993.07a
• /
• pp.535-543
• /
• 1993

이 논문의 목적은 기존의 상수도 관망을 개선하는데 있어 어느 관을 교체 또는 재생할 것인가와 펌프용량을 얼마나 늘릴 것인가를 결정함으로써 관망내 각 급수지점에서의 요구유량및 수압을 만족시킴은 물론 그에 드는 비용을 최소화시키는 모델을 개발하는데 있다. 이 논문은 관교체 비용, 세관 및 재생 비용, 관보수 비용, 펌핑 비용, 펌프시설 확충비용 등의 다섯가지 비용들을 비교 검토함으로써 의사결정을 하게 된다. 제약조건식으로는 급수 조건식, 에너지 방정식, 수리학적 방정식, 결정 조건식, 한계 조건식, 정수 조건식 등이 있다. 이 모델을 수식화하면 정수혼합 비선형계획법 (mixed-integer nonlinear programming, MINLP) 문제가 된다. 이 문제를 풀기 위해 비선형해법의 GRG (generalized reduced gradient) 방법과 분기와 한계 (banch and bound) 기법을 통한 implicit enumeration 기법을 접목시키는 방법을 제안하였다.

• Journal of the Korean Mathematical Society
• /
• v.59 no.3
• /
• pp.519-548
• /
• 2022

In this paper, a stabilized-penalized collocated finite volume (SPCFV) scheme is developed and studied for the stationary generalized Navier-Stokes equations with mixed Dirichlet-traction boundary conditions modelling an incompressible biological fluid flow. This method is based on the lowest order approximation (piecewise constants) for both velocity and pressure unknowns. The stabilization-penalization is performed by adding discrete pressure terms to the approximate formulation. These simultaneously involve discrete jump pressures through the interior volume-boundaries and discrete pressures of volumes on the domain boundary. Stability, existence and uniqueness of discrete solutions are established. Moreover, a convergence analysis of the nonlinear solver is also provided. Numerical results from model tests are performed to demonstrate the stability, optimal convergence in the usual L² and discrete H¹ norms as well as robustness of the proposed scheme with respect to the choice of the given traction vector.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.30 no.4B
• /
• pp.399-412
• /
• 2010

The goal of this research is to apply the neural networks models for the temporal disaggregation of the yearly pan evaporation (PE) data, Republic of Korea. The neural networks models consist of multilayer perceptron neural networks model (MLP-NNM) and generalized regression neural networks model (GRNNM), respectively. And, for the performances evaluation of the neural networks models, they are composed of training and test performances, respectively. The three types of data such as the historic, the generated, and the mixed data are used for the training performance. The only historic data, however, is used for the testing performance. From this research, we evaluate the application of MLP-NNM and GRNNM for the temporal disaggregation of nonlinear time series data. We should, furthermore, construct the credible monthly PE data from the temporal disaggregation of the yearly PE data, and can suggest the available data for the evaluation of irrigation and drainage networks system.

• The Korean Journal of Applied Statistics
• /
• v.34 no.5
• /
• pp.697-710
• /
• 2021

Relative error prediction is preferred over ordinary prediction methods when relative/percentile errors are regarded as important, especially in econometrics, software engineering and government official statistics. The relative error prediction techniques have been developed in linear/nonlinear regression, nonparametric regression using kernel regression smoother, and stationary time series models. However, random effect models have not been used in relative error prediction. The purpose of this article is to extend relative error prediction to some of generalized linear mixed model (GLMM) with panel data, which is the random effect models based on gamma, lognormal, or inverse gaussian distribution. For better understanding, the real auto insurance data is used to predict the claim size, and the best predictor and the best relative error predictor are comparatively illustrated.