• 한국시스템다이내믹스연구
• /
• 제14권3호
• /
• pp.105-134
• /
• 2013

Electrical construction business has public and professional characters. It may require appropriate interventions of the government because these business activities stand for not only profit-seeking competition, but also supplies of one of the key functions in our society. In other words, public benefit and private benefit are still in existence. The government therefore considers such an aspect of public importance of the business sector and needs to plan to adjust technical and engineering manpower of this market. This study focuses on the imbalance for labor supply and demand of technical engineer in electrical construction business. A system dynamics analysis is applied to understand and simulate the imbalance as a soft approach. It has the merit of causal loop diagram to alleviate the limitation of data lack problem. We find that excess demand is expected from 2010 to 2011, and excess supply is predicted from 2012 to 2021 about the manpower of technical engineer. It shows considerable disagreement between the supply and demand of human resource. So we suggest that it is strong necessity to construct statistics infrastructure for a manpower supply and demand plan.

• Safety and Health at Work
• /
• 제13권4호
• /
• pp.469-474
• /
• 2022

Background: Nowadays, counter-productive work behaviors (CWBs) have turned into a common and costly position for many organizations and especially health centers. Therefore, the study was carried out to examine and compare the demand-control-support (DCS) and effort-reward imbalance (ERI) models as predictors of CWBs. Methods: The study was cross-sectional. The population was all nurses working in public hospitals in Hamadan, Iran of whom 320 were selected as the sample based on simple random sampling method. The instruments used were Job Content Questionnaire, Effort-Reward Imbalance Questionnaire, and Counterproductivity Work Behavior Questionnaire. Data were analyzed using correlation and regression analysis in SPSS18. Results: The findings indicated that both ERI and DCS models could predict CWB (p ≤ 0.05); however, the DCS model variables can explain the variance of CWB-I and CWB-O approximately 8% more than the ERI model variables and have more power in predicting these behaviors in the nursing community. Conclusion: According to the results, job stress is a key factor in the incidence of CWBs among nurses. Considering the importance and impact of each component of ERI and DCS models in the occurrence of CWBs, corrective actions can be taken to reduce their incidence in nurses.

• Journal of Electrical Engineering and Technology
• /
• 제8권6호
• /
• pp.1310-1319
• /
• 2013

Resolvers are transducers that are used to sense the angular position of rotational machines. The analog resolver is necessary to use resolver to digital converter. Among the RDC software method, angle tracking observer (ATO) is the most popular method. In an actual resolver-based position sensing system, amplitude imbalance dominantly distorts the estimate position information of ATO. Minority papers have reported position error compensation of resolver's output signal with amplitude imbalance. This paper proposes new ATO algorithm in order to compensate position errors caused by the amplitude imbalance. There is no need premeasured off line data. This is easy, simple, cost-effective, and able to work on line compensation. To verify feasibility of the proposed algorithm, simulation and experiments are carried out.

• 한국안전학회지
• /
• 제37권5호
• /
• pp.62-73
• /
• 2022

This study investigates the working conditions of shift workers according to age group by analyzing the sixth Korean Working Conditions Survey's data. A total of 1,323 shift workers were extracted from the dataset. Three age groups (A: 20s-30s, B: 40s-50s, C: 60s and above) were statistically compared in terms of health problems, emotional hazard exposure, work-life imbalance, and satisfaction with working conditions. Elderly shift workers (those in their 60s and above) had significantly more severe health problems and work-life imbalance, greater exposure to emotional hazards, and lower satisfaction with working conditions than young shift workers (those in their 20s-50s). The study's findings reveal the characteristics of working conditions for elderly shift workers and would be useful for improving shift workers' quality of life, as well as safety and productivity in the workplace.

• 산업융합연구
• /
• 제22권1호
• /
• pp.177-183
• /
• 2024

본 연구는 데이터의 품질이 인공지능(AI) 성능에 미치는 영향을 검토한다. 이를 위해, 데이터 특성변수(Feature)의 유사도와 클래스(Class) 구성의 불균형을 고려한 모의실험(Simulation)을 통해 라벨링 오류 수준이 인공지능의 성능에 미치는 영향을 비교 분석하였다. 그 결과, 특성변수 간 유사성이 높은 데이터에서는 특성 변수 간 유사성이 낮은 데이터에 비해 라벨링 정확도에 더 민감하게 반응하였으며, 클래스 불균형이 증가함에 따라 인공지능 정확도가 급격히 감소되는 경향을 관찰하였다. 이는 인공지능 학습데이터의 품질평가 기준 및 관련 연구를 위한 기초자료가 될 것이다.

• 국제학술발표논문집
• /
• The 9th International Conference on Construction Engineering and Project Management
• /
• pp.808-813
• /
• 2022

This study presents the problems of data imbalance, varying difficulties across target objects, and small objects in construction object segmentation for far-field monitoring and utilize compound loss functions to address it. Construction site scenes of assembling scaffolds were analyzed to test the effectiveness of compound loss functions for five construction object classes---workers, hardhats, harnesses, straps, hooks. The challenging problem was mitigated by employing a focal and Jaccard loss terms in the original loss function of LinkNet segmentation model. The findings indicates the importance of the loss function design for model performance on construction site scenes for far-field monitoring.

• Journal of Chest Surgery
• /
• 제49권sup1호
• /
• pp.14-19
• /
• 2016

Background: This study uses the relevance index to understand the condition of regional medical service use for cardiovascular surgery and to identify the medical service use imbalance between regions. Methods: This study calculated the relevance index of 16 metropolitan cities and provinces using resident registration address data from the Ministry of Government Administration and Home Affairs and the 2010-2014 health insurance, medical care assistance, and medical benefits claims data from the Health Insurance Review and Assessment Service. We identified developments over the 5-year time period and analyzed the level of regional imbalance regarding cardiovascular surgery through the relative comparison of relevance indexes between cardiovascular and other types of surgery. Results: The relevance index was high in large cities such as Seoul, Daegu, and Gwangju, but low in regions that were geographically far from the capital area, such as the Gangwon and Jeju areas. Relevance indexes also fell as the years passed. Cardiovascular surgery has a relatively low relevance index compared to key types of surgery of other fields, such as neurosurgery and colorectal surgery. Conclusion: This study identified medical service use imbalance between regions for cardiovascular surgery. Results of this study demonstrate the need for political intervention to enhance the accessibility of necessary special treatment, such as cardiovascular surgery.

• 스마트미디어저널
• /
• 제8권2호
• /
• pp.46-57
• /
• 2019

In recent research, deep learning-based methods have achieved state-of-the-art performance in various computer vision tasks. However, these methods are commonly supervised, and require huge amounts of annotated data to train. Acquisition of data demands an additional costly effort, particularly for the tasks where it becomes challenging to obtain large amounts of data considering the time constraints and the requirement of professional human diligence. In this paper, we present a data level synthetic sampling solution to learn from small and imbalanced data sets using Generative Adversarial Networks (GANs). The reason for using GANs are the challenges posed in various fields to manage with the small datasets and fluctuating amounts of samples per class. As a result, we present an approach that can improve learning with respect to data distributions, reducing the partiality introduced by class imbalance and hence shifting the classification decision boundary towards more accurate results. Our novel method is demonstrated on a small dataset of 2789 tomato plant disease images, highly corrupted with class imbalance in 9 disease categories. Moreover, we evaluate our results in terms of different metrics and compare the quality of these results for distinct classes.

• Safety and Health at Work
• /
• 제10권2호
• /
• pp.205-212
• /
• 2019

Background: Workforce health is one of the primary and most challenging issues, particularly in industrialized countries. This article aims at modeling the major factors affecting accidents in the workplace, including general health, work-family conflict, effort-reward imbalance, and internal and external locus of control. Methods: A cross-sectional study was conducted in Esfahan Steel Company in Iran. A total of 450 participants were divided into two groups-control and case-and the questionnaires were distributed among them. Data were collected through a 7-part questionnaire. Finally, the results were analyzed using SPSS 22.0 and Amos software. Results: All the studied variables had a significant relationship with the accident proneness. In the case group, general health with a coefficient of -0.37, worke-family conflict with 0.10, effort-reward imbalance with 0.10, internal locus of control with -0.07, and external locus of control with 0.40 had a direct effect on occupational stress. Occupational stress also had a positive direct effect on accident proneness with a coefficient of 0.47. In addition, fitness indices of control group showed general health (-0.35), worke-family conflict (0.36), effort-reward imbalance (0.13), internal locus of control (-0.15), and external locus of control (0.12) have a direct effect on occupational stress. Besides, occupational stress with a coefficient of 0.09 had a direct effect on accident proneness. Conclusion: It can be concluded that although previous studies and the present study showed the effect of stress on accident and accident proneness, some hidden and external factors such as work-family conflict, effort-reward imbalance, and external locus of control that affect stress should also be considered. It helps industries face less occupational stress and, consequently, less occurrence rates of accidents.

• 한국콘텐츠학회논문지
• /
• 제18권8호
• /
• pp.525-535
• /
• 2018

분류 문제에서 특정 범주의 빈도가 다른 범주에 비해 과도하게 높은 경우, 왜곡된 기계 학습을 유발할 수 있는 데이터 불균형(imbalanced data) 문제가 발생한다. 기업부도 예측 문제도 그 중 하나인데, 일반적으로 금융기관과 거래하는 기업들의 부도율은 대단히 낮아서, 부도 사례보다 정상 사례의 빈도가 월등히 높은 데이터 불균형 문제가 발생하고 있다. 이러한 데이터 불균형 문제를 해결하기 위해서는 적절한 표본추출 기법이 적용될 필요가 있으며, 지금껏 소수 범주 데이터를 복원 추출함으로써 다수 범주 데이터와 비율을 맞추어 데이터 불균형을 해결하는 오버 샘플링(oversampling) 기법이 주로 활용되어 왔다. 그러나 전통적인 오버 샘플링은 과적합화(overfitting)가 발생할 위험이 높아질 수 있는 단점이 있다. 이러한 배경에서 본 연구는 효과적인 기업부도 예측 모형 학습을 위한 표본추출 기법으로 2014년에 Menardi와 Torelli가 제안한 ROSE(random over sampling examples) 기법을 제안한다. ROSE 기법은 학습에 사용될 사례를 반복적으로 새롭게 합성하여 생성(synthetic generation)하는 기법으로, 과적합화 문제를 회피하면서도 분류 예측 정확도 개선에 도움을 줄 수 있다. 이에 본 연구에서는 ROSE 기법을 가장 성능이 우수한 이분류기로 알려진 SVM(support vector machine)과 결합하여 국내 한 대형 은행의 기업부도 예측에 적용해 보고, 다른 표본추출 기법들과의 비교연구를 수행하였다. 실험 결과, ROSE 기법이 다른 기법에 비해 통계적으로 유의한 수준으로 SVM의 예측정확도 개선에 기여할 수 있음을 확인하였다. 이러한 본 연구의 결과는 부도예측 외에 다른 사회과학 분야 예측문제의 데이터 불균형 문제 해결에도 ROSE가 우수한 대안이 될 수 있다는 사실을 시사한다.