• The Journal of the Korea Contents Association
• /
• v.21 no.11
• /
• pp.135-144
• /
• 2021

Recently, data-driven decision-making technology has become a key technology leading the data industry, and machine learning technology for this requires high-quality training datasets. However, real-world data contains missing values for various reasons, which degrades the performance of prediction models learned from the poor training data. Therefore, in order to build a high-performance model from real-world datasets, many studies on automatically imputing missing values in initial training data have been actively conducted. Many of conventional machine learning-based imputation techniques for handling missing data involve very time-consuming and cumbersome work because they are applied only to numeric type of columns or create individual predictive models for each columns. Therefore, this paper proposes a new data imputation technique called 'Denoising Self-Attention Network (DSAN)', which can be applied to mixed-type dataset containing both numerical and categorical columns. DSAN can learn robust feature expression vectors by combining self-attention and denoising techniques, and can automatically interpolate multiple missing variables in parallel through multi-task learning. To verify the validity of the proposed technique, data imputation experiments has been performed after arbitrarily generating missing values for several mixed-type training data. Then we show the validity of the proposed technique by comparing the performance of the binary classification models trained on imputed data together with the errors between the original and imputed values.

• Journal of Practical Engineering Education
• /
• v.15 no.3
• /
• pp.671-679
• /
• 2023

This study aims to establish an evaluation model by quantifying the evaluation index as a follow-up study to the development of evaluation index for work-study parallel learning companies. An evaluation model was established by verifying the 2nd level components based on the quantitative factors of the learning company, the qualitative factors, the competency factors of the person in charge, and the competency factors of the learning workers, which are the highest-level components derived from previous study. For the evaluation of a learning company, an AHP survey was conducted with experts in charge of the company consulting to derive important factors that determine the quality of on-site education and training, and the evaluation model of the learning company was completed and grouped by calculating the weight between evaluation items proceeded. Work-study parallel program was promoted as a key policy to resolve the mismatch between industrial sites and school education and realize a competency-centered society, and as of December 2022, 16,664 companies participated in the training. Learning companies play a very important role as education and training supply organizations that conduct field training. It is expected that the support and consulting plan for each level of learning companies according to the evaluation model presented in this study will be used as basic data to improve the quality of work-study parallel program.

• Korea Journal of Hospital Management
• /
• v.6 no.1
• /
• pp.120-146
• /
• 2001

This study was carried out to provide the essential information in improving the graduate medical education in Korea. For the study, a survey targeting the directors of GME of nationwide teaching hospitals was performed with a questionnaire asking the questions such as the director's perception on the quality of GME, trainees' salary level, trainees' specialty selection tendency, training system and its duration. The collected data were analyzed using t-test, ANOVA, and $x^2$-test. The results were as follows: 1. The survey were executed on 240 teaching hospitals in Korea and the response rate was 66.2% (159 hospitals replied). 2. The bigger a hospitals is the better in Quality of education. Larger hospitals tend to have better status in all items including medical specialists' experience, contents of medical curriculum, general environment for medical education and medical trainees's salary level. The result supported the general perception on the positive relationship between hospital size and Quality of GMA. 3. Providing convenience for medical trainees who prepares for the medical specialist Qualifying examination didn't affect the results of the examination. 4. The directions of GME have a perception that the trainees give positive impact on financial performance of their hospitals. This seems to be one of the reasons that hospitals try to retain as many trainees as possible. 5. The directors of GME considered medical trainees as an educate, and most of them responded positively on the need of governmental supports for the education cost and the trainee's salary. Considering above results, it seems that GME would get more social attention and the trainees' impact on hospitals operation would be increased more than before. In response to these trends, hospitals would find out the ways to lower dependency on trainees, and this change of attitude of hospitals on the GME would cause problems in operation of hospitals and GME itself. In order to prevent these problems the policy on GME should be directed in following ways. 1. The contents of Qualifying examination for specialist should be improved. 2. The curriculum of GME should be strictly followed. 3. The status of trainee in a hospital has to be defined as eductee. 4. Government has to support a half of the education cost and salary of trainee. 5. The distribution of the trainee among the hospital group have to be based on total available. 6. The financial support and welfare of trainee should be improved gradually and systematically.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.10
• /
• pp.430-436
• /
• 2018

Special railway vehicles for track maintenance are equipped with a tamping device that adjusts various track trajectories to reduce the vibration of rolling stock and improve ride quality during trains passing over a track. The development of a simulator that can confirm the error of the actual tamping work is important for reducing human error in the linearization of the track misalignment. In this study, to improve the reality and training effect of conventional 2D simulator, 3D simulator modeling was implemented for tamping work of special railway vehicles in virtual space. The problem of buffering during high screen quality of tamping work was solved using the Unwrap UVW mapping technique of a low polygon extracted from high quality polygon modeling. The human error in the training of the tamping work was detected by the principle of circle and square collision when the tamping tyne and the sleeper collided. In addition, vibration of the driving chair was generated at the same time as the collision, and the number of the sleeper strikes is displayed on the simulator exercise screen. Owing to the scattering of railway ballast protruding from the sleepers, which had a serious effect on the safety of the vehicle, the gravel bouncing effect of the tamping unit was applied.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.21 no.3
• /
• pp.223-229
• /
• 2020

The current technological advances are changing the way people live. In the history of war in the past, nations or people with an edge in military science and technology have always been the victor. The emergence of new technologies not only alters war patterns drastically but also affects military operational concepts and organizational systems. As the paradigm of the security environment changes, developed countries are investing heavily in defense R & D for their security. Therefore, the importance of professionalism of the defense acquisition personnel who perform their tasks from the exploration and development of weapon systems to mass production, deployment, and logistics support cannot be overemphasized. In the United States, an advanced country in this field, to improve the work efficiency of acquisition personnel in the defense field, the Directive of Ministry of Defense issued a law on education and training career development for personnel in 1990. The present study refers to related materials, such as the Desk Guide for the main education requirements for the US acquisition personnel in the military field, which contributes to the improvement of the workforce of the acquired manpower after the systematic education system.

• Horticultural Science & Technology
• /
• v.18 no.4
• /
• pp.503-507
• /
• 2000

This study was conducted to investigate the effect of shoot training method on the plant growth and fruit quality and yield of 'Sinsakigake-2' and 'Shishito' peppers (Capsicum annuum L.) grown in the glasshouse. Plants were either left untrained as control or trained at the third node leaving two or four shoots per plant. The untrained control plants had no pruning and therefore had all the lateral branches. The growth was enhanced in plants with two trained shoots in both cultivars. Fruit length and width, fruit weight, and pericarp thickness were not affected by the number of shoots trained. However, the percent marketable fruits was the highest in plants with two trained shoots, and the number of marketable fruits per plant was the highest in plants with four trained shoots. Marketable yield in plants with four trained shoots increased 15% in 'Sinsakigake-2' and 5% in 'Shishito' as compared to that of the control. Results of this study showed that yield and quality of pepper fruits were promoted by training with four shoots and the effect was more pronounced in 'Sinsakigake-2' than 'Shishito' pepper.

• Journal of the Korean Society of Radiology
• /
• v.14 no.7
• /
• pp.991-1001
• /
• 2020

Supervised deep learning technologies for improving the image quality of computed tomography (CT) need a lot of training data. When input images have different characteristics with training images, the technologies cause structural distortion in output images. In this study, an imaging model based on the deep reinforcement learning (DRL) was developed for overcoming the drawbacks of the supervised deep learning technologies and reducing noise in CT images. The DRL model was consisted of shared, value and policy networks, and the networks included convolutional layers, rectified linear unit (ReLU), dilation factors and gate rotation unit (GRU) in order to extract noise features from CT images and improve the performance of the DRL model. Also, the quality of the CT images obtained by using the DRL model was compared to that obtained by using the supervised deep learning model. The results showed that the image accuracy for the DRL model was higher than that for the supervised deep learning model, and the image noise for the DRL model was smaller than that for the supervised deep learning model. Also, the DRL model reduced the noise of the CT images, which had different characteristics with training images. Therefore, the DRL model is able to reduce image noise as well as maintain the structural information of CT images.

• Journal of the Korea institute for structural maintenance and inspection
• /
• v.28 no.4
• /
• pp.55-61
• /
• 2024

Recently, the number of aging concrete structures is steadily increasing. This is because many of these structures are reaching their expected lifespan. Such structures require accurate inspections and persistent maintenance. Otherwise, their original functions and performance may degrade, potentially leading to safety accidents. Therefore, research on objective inspection technologies using deep learning and computer vision is actively being conducted. High-resolution images can accurately observe not only micro cracks but also spalling and exposed rebar, and deep learning enables automated detection. High detection performance in deep learning is only guaranteed with diverse and numerous training datasets. However, surface damage to concrete is not commonly captured in images, resulting in a lack of training data. To overcome this limitation, this study proposed a method for generating concrete surface damage images, including cracks, spalling, and exposed rebar, using stable diffusion. This method synthesizes new damage images by paired text and image data. For this purpose, a training dataset of 678 images was secured, and fine-tuning was performed through low-rank adaptation. The quality of the generated images was compared according to three base models of stable diffusion. As a result, a method to synthesize the most diverse and high-quality concrete damage images was developed. This research is expected to address the issue of data scarcity and contribute to improving the accuracy of deep learning-based damage detection algorithms in the future.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.23 no.3
• /
• pp.243-249
• /
• 2013

Objectives: This study assessed the status of domestic industrial hygiene laboratories using data from on-site investigation for revision of quality control systems in 2012-2013. Methods: The target laboratories were 60 industrial hygiene laboratories chosen by random selection and nationwide distribution which had participated in on-site investigations for revision of quality control systems from March 2012 to August 2013. The investigation was performed on-site following standard quality control procedures. The score between each group was compared using Mann-Whitney and Kruskal-Wallis tests, and the correlation between analytical career, sex, academic major of analyst and score of analytical performance was expressed as Spearman's rank correlation coefficient. Results: The assessment revealed that the items to be improved, in sequence, were effort at staff training (score 65.5), ability to calculate data (score 73.4), establishment of internal quality control guidelines (score 75.7), laboratory facilities (score 77.1), degree of understanding and skill at gas chromatography (score 77.1). Analysis performance showed a positive correlation with career of analyst (r=0.56, p<0.01). Conclusions: The practice of on-site investigation for quality control systems showed the current status of industrial hygiene laboratories in the first trial. There were many laboratories which needed improvement and development of analytical systems. This assessment can provide information for the systematic operation and improvement of facilities at each laboratory. Further practice of this investigation will lead to a proficiency testing and accreditation system for autonomous quality control as is the practice in many countries, rather than mandatory practice by legal regulation.

• Quality Improvement in Health Care
• /
• v.21 no.2
• /
• pp.40-56
• /
• 2015

Objectives: To outline overall duties of quality improvement (QI) performers within a health care organization, thus describing their key tasks, including task element-related frequency, importance and difficulty in enough detail. Methods: A DACUM (Developing A CurriculUM) workshop took place to outline overall job activities of QI performers. To examine the scope of their duty and task, we performed a questionnaire survey of 338 QI performers from 111 hospitals. Results: The results of our survey showed that for the task assigned to each QI performer, there were 10 duties, 31 tasks and 119 task elements. Respondents cited a project planning as the most frequent/important duty, and a research was the highest level of difficulty in their duty. They also said that the most frequent task was index management, the most important task was a business plan, and the highest level of difficulty was a practical application of QI research. QI performers added that the most frequent task element was receipt of patient safety reporting in patient safety system, the most important task element was an analysis for patient safety and its improvement, and the highest level of difficulty was a regional influence analysis related to the patient safety and its improvement. Conclusion: To ensure that QI performers play a pivotal role as a manager to better improve patient safety and the quality of health care services, proper training program for them should be developed by reflecting the results of our study.