• The Korean Journal of Applied Statistics
• /
• v.30 no.5
• /
• pp.793-808
• /
• 2017

We compare various variable screening methods on multiclass classification problems when the data is ultrahigh-dimensional. Two different approaches were considered: (1) pairwise extension from binary classification via one versus one or one versus rest comparisons and (2) direct classification of multiclass responses. We conducted extensive simulation studies under different conditions: heavy tailed explanatory variables, correlated signal and noise variables, correlated joint distributions but uncorrelated marginals, and unbalanced response variables. We then analyzed real data to examine the performance of the methods. The results showed that model-free methods perform better for multiclass classification problems as well as binary ones.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.2
• /
• pp.191-197
• /
• 2021

This study purports to explore potential determinants of welfare attitudes toward universalism vs selectivism. For this purpose, literature review upon such subjects as definitions of universalism and selectivism and welfare attitudes has been done. The hierarchical regression analyses show several major results. First and foremost, the effects of those variables such as political orientation and attitudes toward free education and gratuitous child care, categorized as political-social stance were found to be significant. However, it was unexpected results that those variables which have been found signigicant in predicting welfare attitudes in previous literature, that is to say age, education and economic status especially were not to be found significant in predicting welfare attitudes toward universalism vs selectivism. There could be many underlying causes for this result including measurement errors, and this study strongly speculates that the division between universalism vs selectivism itself exists only both in purely conceptual level and in political rhetoric and therefore, universalism or selectivism as people's consistent and logical attitudes or consciousness may simply not exist at all.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.22 no.5
• /
• pp.110-118
• /
• 2018

Due to recent earthquakes, there is a growing awareness that Korea is not a safe zone for earthquakes any more. Therefore, the review of various aspects of the seismic safety of the infrastructures are being carried out. Because of the characteristics of the underground structure buried in the ground, the electric power utility tunnels must be considered not only for the inertia and load capacity of the structure itself but also the characteristics of the surrounding soils. An extensive and accurate numerical analysis is inevitably required in order to consider the interaction with the ground, but it is difficult to apply the soil-structure interaction analyses, which generally requires high cost and extensive time, to all electric power utility tunnel structures. In this study, the major design variables including soil characteristics are considered as independent variables, and the seismic safety factor, which is the result of the numerical analysis, is considered as a dependent variable. Thus, a method is proposed to select vulnerable electric power utility tunnels with low seismic safety factor while excluding costly and time-consuming numerical analyses through the direct correlation analysis between independent and dependent variables. Equations of boundary limits were derived based on the distribution of the seismic safety factor and the cover depth and rebar amounts with high correlation relationship. Consequently, a very efficient and simple approach is proposed to select vulnerable electric power utility tunnels without intensive numerical analyses. Among the 108 electric power utility tunnels that were investigated in this paper, 30% were screened as fragile structures, and it is confirmed that the screening method is valid by checking the safety factors of the fragile structure. The approach is relatively very simple to use and easy to expand, and can be conveniently applied to additional data to be obtained in the future.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2016.05a
• /
• pp.151-151
• /
• 2016

하천 유역의 수자원관리에 있어서 홍수 및 가뭄 기간에 유출의 규모와 빈도와 같은 유출특성을 신뢰할 수 있도록 예측하는 것은 매우 중요하다. 수문모형은 이러한 유역의 신뢰성 있는 유출량 예측을 위해 이용되며, 수문모형의 결과물은 수문순환 과정의 공간적 표출이나 매개변수 추정방법 등 다양한 요인에 매우 민감하게 반영된다. 대부분의 수문모형 매개변수들은 해당 유역의 특성이나, 홍수 및 가뭄과 같은 극단적 유출상황에 따라 설정되어 있지 못한 실정이며, 이는 모형의 신뢰성 있는 보정 및 유출량 모의를 보다 정밀하게 수행하지 못하는 원인으로 작용하게 된다. 본 연구의 목적은 SWAT(Soil and Water Assessment Tool) 모형을 이용하여 유역의 하천 유출량을 모의함에 있어서 홍수 및 가뭄년, 평년의 유출특성이 모형의 매개변수 추정에 미치는 영향을 분석하고자 하는 것이다. 이를 위해, 안성천 유역($1,658.7km^2$)을 대상으로 유역 내 3지점의 기상관측소(이천, 수원, 천안)를 대상으로 40년(1976~2015)동안의 일 기상자료를 수집하여 SWAT 모형을 구축하였다. 홍수년 및 가뭄년, 평년을 포함하는 선별된 기간에 대하여 다양한 목적함수($R^2$, NSE, RMSE, PBIAS)를 활용하여 각각의 조합된 기간의 극단적 유출특성에 초점을 맞추어 검보정을 수행하였다. 이후 홍수년 및 가뭄년, 평년을 포함하는 선별된 기간에서의 유출량의 규모와 빈도에 영향을 미치는 매개변수를 도출하고 민감도를 평가하였다.

• Resources Recycling
• /
• v.23 no.6
• /
• pp.30-39
• /
• 2014

The trajectories of PCB(Printed Circuit Board) particles in the corona discharge electrostatic separation was simulated. The PCB particles are prepared by crushing bare board, which disassembled from electronic components, consist mostly of copper and FR-4(Flame Retardant Level-4) Firstly, a model was established for calculating of detachment points of PCB particles from the rotating electrode in separator. The model of detachment points was derived from equilibrium of force such as gravity force, centrifugal force, electrostatic force. The trajectories of particles after detachment was calculated by acceleration derived from time-integrating method of motion equation. In this simulation, particle size, supplied voltage, rotation speed of rotating roll electrode and angle of induction electrode were adopted as variables. While the trajectories of FR-4 particles were affected by all variables, rotation speed of rotating roll electrode was dominant variables affecting trajectories of copper particles.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.8
• /
• pp.379-388
• /
• 2016

There are various system defects from weapons manufacturing due to the numerous production processes and various production environments. The first kind of defect is patent defects, which can be detected by visual inspection, functional testing, and existing quality control procedures during the manufacturing process. The second kind is latent defects, which cannot be detected though existing quality management approaches because of the complexity of the system and manufacturing process. To minimize the initial defect problems, environmental stress screening (ESS) is needed to detect the defects, remove them, and improve the product conditions based on the environmental stress conditions of temperature and vibration. We implemented a tool for quantitative ESS effectiveness analysis and profile design automation based on MIL-HDBK-344 and verified it using six scenarios with different temperature stress, vibration stress, and test designs.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.3
• /
• pp.227-231
• /
• 2024

This study purports to examine the potential effects of welfare attitudes of Korean people upon their political orientation. The 17^th Korea Welfare Panel Data(KWPD) in 2022 are used for this purpose. Independent variable include sex, age, education, interest in politics, and employment status. Discriminant analysis show several results. First and foremost, pre-established discriminant function works well for classification of respondents' liberal vs conservative stance. Secondly, except gender and dummy variable for temporary employed, all independent variables contribute significantly for the classification at a given significance level. . Finally, welfare attitudes of respondents', measured by universalism vs selectivism and the attitudes upon increasing tax for welfare expenditures are found to be significant and relatively big impacts upon dependent variable, compard to other variables in the model. The nature of causal relationship between welfare attitudes and political orientation remains for further study.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.1
• /
• pp.181-187
• /
• 2016

Silica is an essential material in the electronics industries of LCDs and OLEDs, which particularly require high purity. This study attempted to find the optimal design of a magnetic separator for silica sand containing iron compounds using CFD simulation. Three designs of magnetic separation were prepared and their efficiency was examined. As a result of the evaluation, the sufficient contact of particulate silica with the surface of magnetic emitters improved the magnetic separation effects. In addition, the loss of $SiO_2$ and the removal rate of $Fe_2O_3$ depended strongly on the particle size, flow rate and magnetic flux density. In addition, magnetic separation is quite effective for a particle size of $10{\mu}m$ with a 0.2 m/s flow rate.

• Journal of the Korean Society for Library and Information Science
• /
• v.34 no.4
• /
• pp.161-174
• /
• 2000

In this conjoint analysis, the objectives of the research are to derive structural part-worth level estimations by the conjoint factors and individual levels from the university library functional factors. The major findings of this study are; (1) The level of temperature of the library is statistically the most important one for the research samples, and (2) Facility factors and its levels as well as operation factors and its levels are statistically shown as less important variables.

• Journal of Korea Multimedia Society
• /
• v.7 no.7
• /
• pp.877-885
• /
• 2004

In this paper, we propose an enhancement shot change detection method using the neural net and the robust feature selection out of multiple features. The previous shot change detection methods usually used single feature and fixed threshold between consecutive frames. However, contents such as color, shape, background, and texture change simultaneously at shot change points in a video sequence. Therefore, in this paper, we detect the shot changes effectively using robust features, which are supplementary each other, rather than using single feature. In this paper, we use the typical CART (classification and regression tree) of data mining method to select the robust features, and the backpropagation neural net to determine the threshold of the each selected features. And to evaluation the performance of the robust feature selection, we compare the proposed method to the PCA(principal component analysis) method of the typical feature selection. According to the experimental result. it was revealed that the performance of our method had better that than the PCA method.