• Survey Research
• /
• v.9 no.2
• /
• pp.1-27
• /
• 2008

Problems of incomplete data are pervasive in statistical analysis. In particular, incomplete data have been an important challenge in repeated measures studies. The objective of this study is to give a brief introduction to missing data mechanisms and conventional/recent missing data methods and to assess the performance of various missing data methods under ignorable and non-ignorable missingness mechanisms. Given the inadequate attention to longitudinal studies with missing data, this study applied recent advances in missing data methods to repeated measures models and investigated the performance of various missing data methods, such as FIML (Full Information Maximum Likelihood Estimation) and MICE(Multivariate Imputation by Chained Equations), under MCAR, MAR, and MNAR mechanisms. Overall, the results showed that listwise deletion and mean imputation performed poorly compared to other recommended missing data procedures. The better performance of EM, FIML, and MICE was more noticeable under MAR compared to MCAR. With the non-ignorable missing data, this study showed that missing data methods did not perform well. In particular, this problem was noticeable in slope-related estimates. Therefore, this study suggests that if missing data are suspected to be non-ignorable, developmental research may underestimate true rates of change over the life course. This study also suggests that bias from non-ignorable missing data can be substantially reduced by considering rich information from variables related to missingness.

• Korean Journal of Culture and Social Issue
• /
• v.17 no.1
• /
• pp.19-49
• /
• 2011

In this study, multilateral conceptualizations of underemployment were measured in terms of wages, social status, skill utilization and permanence of the job, and then the effects of antecedents on underemployment and the effects of underemployment on organizational adaptation were examined. Data obtained by a longitudinally designed survey at intervals of 18 months with the reemployed(N = 153) after job loss were used. The underemployment measures include 1) the ratio of wage change 2) the ratio of status change 3) the ratio of education 4) the occurrence of change from the permanent job to temporary job, 5) overqualification - growth opportunity, 6) overqualification - mismatch. The first four measures are social-economic and objective measures and the last two measures are psychological and self-reported ones. Demographic variables(sex, age, education level, and period of unemployed), circumstantial variables(economic hardship, number of dependents), and psychological variables(job-seeking self-efficacy, depression/anxiety, latent function) are included in antecedents. In the effects of antecedents on underemployment, age increases the level of underemployment in the aspects of wage and job status. Economic hardship increases the possibility of underemployment in the aspects of education and number of dependents increases the possibility of underemployment in the aspects of job status. Job seeking self-efficacy decreases the possibility of underemployment in the overqualification - no growth. Retention of latent function during the period of unemployment lowers the possibility of underemployment in the overqualification - no growth. The level of depression and anxiety during the period of unemployment raises the possibility of underemployment in terms of education and in the overqualification - mismatch. In the effects of underemployment on organizational adaptation, the higher the level of underemployment in the aspect of education is, the lower the level of person-organization fit, emotional commitment, and job satisfaction are. And the transition from permanent job to temporary job makes emotional commitment and job satisfaction lower. No growth and mismatch exerted a significant influence on organizational adaptation generally.

• Journal of the Korea Convergence Society
• /
• v.11 no.3
• /
• pp.1-7
• /
• 2020

This paper introduces two different techniques for dealing with incomplete data and algorithms for learning this data. The first method is to process the incomplete data by assigning the missing value with equal probability that the missing variable can have, and learn this data with the SVM. This technique ensures that the higher the frequency of missing for any variable, the higher the entropy so that it is not selected in the decision tree. This method is characterized by ignoring all remaining information in the missing variable and assigning a new value. On the other hand, the new method is to calculate the entropy probability from the remaining information except the missing value and use it as an estimate of the missing variable. In other words, using a lot of information that is not lost from incomplete learning data to recover some missing information and learn using deep learning. These two methods measure performance by selecting one variable in turn from the training data and iteratively comparing the results of different measurements with varying proportions of data lost in the variable.

• Journal of Korea Multimedia Society
• /
• v.3 no.5
• /
• pp.550-556
• /
• 2000

This paper comes to derive optimal decision rule from incomplete information using the concept of indiscernibility relation and approximation space in Rough set. As there may be some errors in case that processing information contains multiple or missing data, the method of removing or minimizing these data is required. Entropy which is used to measure uncertainty or quantity in information processing field is utilized to remove the incomplete information of rough relation database. But this paper does not always deal with the information system which may be contained incomplete information. This paper is proposed object relation entropy and attribute relation entropy using Rough set as information theoretical measures in order to remove the incomplete information which may contain condition attribute and decision attribute of information system.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.15 no.5
• /
• pp.1426-1432
• /
• 1991

본 연구에서는 Timoshenko보이론과 전달행렬법에 의해서 얻어지는 베어링의 반력들과 주축계의 불완전진동모드간의 선형관계를 이용하여 원리적으로 반복계산없이 반경방향과 굽힘모멘트방향의 선형등방성 베어링계수들을 규명하는 방법을 제안하였다. 제안된 규명방법은 주축계에 사용된 베어링의 수보다 진동모드에 대한 측정점의 수가 2배 이상 많아야 한다는 조건을 가지고 있다. 또한 선형 연립방정식으로 부터 직접 규명된 베어링계수들의 일부가 정보의 부정확성에 의해서 물리적으로 타당성이 없는 음수로 나타나는 경우에는 측정결과와의 차이를 최소화시키면서, 물리적으로 타당성이 있는 양수의 베어링계수들을 규명할 수 있는 방법이 제안조건들을 가진 최적화문제의 형태로 제시되었다. 그리고 제시된 최적화문제의 해는 선형화방법(linearization method)를 통해서 얻었다. 아울러서 주축계의 실험모델에 대한 가진실험결과를 이용 하여 제안된 규명방법의 유용성을 평가하였다.

• Journal of Korean Society of Transportation
• /
• v.30 no.5
• /
• pp.83-90
• /
• 2012

Driver ordinarily takes sudden braking when urgent situation is developed or when the vehicle is involved in an unexpected accident. Therefore, the most common trace at a traffic accident scene is skid mark. Currently, in investigating traffic accident, overspeed is determined by the length of skid mark. However, in order to identify accurate cause of accident, estimation of pre-braking speed which takes into account speed reduction during transient time should be considered as a requirement. In a recent study, several ways to estimate pre-braking speed were suggested, but none considered to differentiate the decelerating transient brake section and skidding section. This study analyzed trends of decelerating transient brake section and skidding section by real braking test.

• Journal of the Earthquake Engineering Society of Korea
• /
• v.8 no.6 s.40
• /
• pp.73-80
• /
• 2004

A nonlinear time-domain system identification algorithm using measured acceleration data is developed for structural damage assessment. To take account of nonlinear behavior of structural systems, an output error between measured and computed acceleration increments has been defined and a constrained nonlinear optimization problem is solved for optimal structural parameters. The algorithm estimates time-varying properties of stiffness and damping parameters. Nonlinear response of restoring force of a structural system is recovered by using the estimated time-varying structural properties and computed displacement by Newmark-$\beta$ method. In the recovery, no pre-defined model for inelastic behavior has been assumed. In developing the algorithm, noise and incomplete measurement in space and state have been considered. To examine the developed algorithm, numerical simulation and laboratory experimental studies on a three-story shear building have been carried out.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.36 no.3
• /
• pp.471-491
• /
• 2016

Traditional limit equilibrium method needs an assumption of the failure surface to calculate the bearing capapcity of the shallow foundation. From the viewpoint of the mechanics of granular materials, however, the failure of the soil mass is initated by the local buckling of the contact force chains. In this study we observed the directional distribution of the contact force chains in the granular assembly stacked by model particles subjected to the model shallow foundation during loading. Two sets of the assemblies with a regular structure and initially local imperfection were prepared for tests. Existence of the initial local imperfection has a significant effect on the directional distribution of the contact force chains. The bearing capacity of the assembly with local imperfection is only 67% the capacity of the assembly with the regular structure.

• Proceedings of the Korea Air Pollution Research Association Conference
• /
• 2001.11a
• /
• pp.358-359
• /
• 2001

PAHs는 불완전연소 과정과 주유소나 타르 생산공장 주변에서 발생하는데 돌연변이와 발암의 잠재력으로 인하여 특별한 관심이 요구되는 것들이고 US-EPA에서도 우선 오염물질로 선정되었다. PAHs를 검출하는 훌륭한 기술들이 존재하나, 이러한 기술들은 연속적인 모니터링이 요구되거나 많은 수의 샘플을 분석해야할 경우에 상당한 양의 비용과 시간이 요구된다. 그러나 분광학 기술을 바탕으로 한 Laser-Induced Fluorescence(LIF) spectroscopy를 이용한 방법은 실시간으로 연속적인 측정을 가능하게 할 뿐 아니라 여러 가지 PAHs, 유류오염물질 등의 물질을 동시에 측정할 수 있다. (중략)

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.22-24
• /
• 2001

정보검색 시스템에서 사용자의 질의어가 불완전함에 따라 생기는 검색 효율의 저하를 줄이기 위하여 용어의 상호관련성을 반영함과 동시에 벡터의 공간을 축소하는 LSI 모델을 사용하여 문서 집합으로부터 잠재적 의미 공간을 구축하였다. 또한 의미 공간상에 있는 문서의 분포에 따라 \"개념\"을 추출하기 하기 위해 k-means algorithm을 사용하여 군집화 시켰다. 이로부터 불완전한 초기 사용자 질의어를 의미 공간에 구축된 클러스터링 정보로 수정하여 새로운 질의어를 생성함으로 검색의 효율을 높이고자 하였다. 검색 효율을 측정하기 위해 TREC 데이터를 이용하여 분석하였으며 결과는 질의어의 성격에 따라 달라졌으나 대체적으로 우수한 성능을 보였다.한 성능을 보였다.