• International journal of advanced smart convergence
• /
• v.13 no.3
• /
• pp.143-149
• /
• 2024

As the size of data and models in machine learning training continues to grow, training on a single server is becoming increasingly challenging. Consequently, the importance of distributed machine learning, which distributes computational loads across multiple machines, is becoming more prominent. However, several unresolved issues remain regarding the performance enhancement of distributed machine learning, including communication overhead, inter-node synchronization challenges, data imbalance and bias, as well as resource management and scheduling. In this paper, we propose ParamHub, which utilizes orchestration to accelerate training speed. This system monitors the performance of each node after the first iteration and reallocates resources to slow nodes, thereby speeding up the training process. This approach ensures that resources are appropriately allocated to nodes in need, maximizing the overall efficiency of resource utilization and enabling all nodes to perform tasks uniformly, resulting in a faster training speed overall. Furthermore, this method enhances the system's scalability and flexibility, allowing for effective application in clusters of various sizes.

• ETRI Journal
• /
• v.45 no.4
• /
• pp.557-569
• /
• 2023

Traditional deterministic channel modeling is accurate in prediction, but due to its complexity, improving computational efficiency remains a challenge. In an alternative approach, we investigated a multilayer artificial neural network (ANN) to predict large-scale and small-scale channel characteristics in metro tunnels. Simulated high-precision training datasets were obtained by combining measurement campaign with a ray tracing (RT) method in a metro tunnel. Performance on the training data was used to determine the number of hidden layers and neurons of the multilayer ANN. The proposed multilayer ANN performed efficiently (10 s for training; 0.19 ms for prediction), and accurately, with better approximation of the RT data than the single-layer ANN. The root mean square errors (RMSE) of path loss (2.82 dB), root mean square delay spread (0.61 ns), azimuth angle spread (3.06°), and elevation angle spread (1.22°) were impressive. These results demonstrate the superior computing efficiency and model complexity of ANNs.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.22 no.5
• /
• pp.410-416
• /
• 2016

After the Sewol Ferry accident, the importance of maritime safety has been emphasized in Korea. In particular, educational and experience training are not only being conducted for maritime personnel but also in schools and at maritime-related organizations in order to broadly instill maritime safety awareness. Based on SOLAS regulations, safety education for sailors conducted every 10 days passenger boats, and fire-fighting drills and abandon-ship training should be conducted once a month on merchant ships. After the Sewol Ferry accident, the maximum number of trainees was reduced from 40 to 20 in order to improve the effectiveness of these training sessions by requiring all trainees to participate in the actual training. The current training process consists of two steps: textbook-based theoretical training and actual practice. Current training environment provides limited capability from human and facility recourses which limit the numbers of trainee participated and system operation time. By introducing the simulation training, it will improve the trainee skill and performance prior to the on-site training and allow the more effective and rapid progress on actual practice. Therefore, it will be proposed the three-step training method in order to improve the effectiveness on fire-fighting drill in Maritime Safety Education on this study. This study suggests a three step training method that would increase the efficiency of maritime safety education. An image-training step to enhance individual task awareness and equipment usage via simulation techniques after theoretical training has been added. To implement this simulation, a virtual training session will be conducted before actual training, based on knowledge obtained from theoretical training, which is expected to increase the speed with which trainees can adapt during the practical training session. In addition, due to the characteristics of the simulation, repeated training is possible for reaction drills in emergency circumstances and other various scenarios that are difficult to replicate in actual training. The efficiency of training is expected to improve because trainees will have practiced before practical training takes place, which will decrease the time needed for practical training and increase the number of training sessions that can be executed, increasing the efficiency of training overall. This study considers development methods for fire-fighting drill simulations using virtual reality techniques.

• Journal of the Ergonomics Society of Korea
• /
• v.33 no.1
• /
• pp.49-58
• /
• 2014

Objective: The aim of this study is to enhance the efficiency of education and training through application and management of 'Transfer of Training' in nuclear power plants. Background: Despite the sophistication and standardization of job-related skills and techniques of workers, accidents/incidents keep taking place due to human errors and unsafe actions and behaviors, which translates into the necessity to review and examine the effectiveness and influence of education and training on the workers of nuclear power plants. Method/Results: This study drew the factors of 'Transfer of Training' through a review on the preceding studies and document research. In addition, through expert examination, this study explored the expected effects and possibility of application when managing the influencing factors of 'Transfer of Training' in nuclear power plants. And lastly, management priority order for nuclear power plants was drawn through an AHP analysis. Conclusion: Among the 'Transfer of Training' factors, the training design factor was the most important. In addition, the design of the training and transfer and goal setting showed a high degree of importance among the influencing factors. Application: The management of 'Transfer of Training' in nuclear power plants enhances the capability of workers and improves the operational integrity of nuclear power plants.

• Proceedings of the CALSEC Conference
• /
• 1998.10a
• /
• pp.141-148
• /
• 1998

ㆍ Design Creation is More Efficient ㆍ Large Shared Body of Knowledge Created and Maintained...Reducing Training Time, Re-Training, Costs and Design Efforts. ㆍ Electronic Data Easier to Exchange and Integrate. An Overall Increase in Efficiency, Savings and Competitiveness Will be Achieved(omitted)

• Journal of Advanced Marine Engineering and Technology
• /
• v.39 no.10
• /
• pp.1054-1061
• /
• 2015

The International Maritime Organization (IMO) has been regulating emissions by making mandatory the compliance with institutions aimed at protecting air quality such as the Energy Efficiency Design Index (EEDI), Ship Energy Efficiency Management Plan (SEEMP) and Tier III. Under the circumstances, one of the response measures considered to be the most feasible is the replacement of existing marine fuel with Liquefied Natural Gas (LNG). The industry has been preemptively building infrastructure and developing and spreading engine technology to enable the use of LNG-fueled ships. The IMO, in turn, recently adopted the International Code of Safety for Ships Using Gases or Other Low-Flash-Point Fuels (IGF Code) as an institutional measure. Thus, it is required to comply with regulations on safety-related design and systems focused on response against potential risk for LNG-fueled ships, in which low-flash-point fuel is handled in the engine room. Especially, the Standards of Training, Certification and Watchkeeping (STCW) Convention was amended accordingly. It has adopted the qualification and training requirements for seafarers who are to provide service aboard ships subject to the IGF Code exemplified by LNG-fueled ships. The expansion in the use of LNG-fueled ships and relevant facilities in fact is expected to increase demand for talents. Thus, the time is ripe to develop methods to set up appropriate STCW training courses for seafarers who board ships subject to the IGF Code. In this study, the STCW Convention and existing STCW training courses applied to seafarers offering service aboard ships subject to the IGF Code are reviewed. The results were reflected to propose ways to design new STCW training courses needed for ships subject to the IGF Code and to identify and improve insufficiencies of the STCW Convention in relation to the IGF Code.

• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.7
• /
• pp.759-768
• /
• 2018

As the number of exchange in various fields between countries increases, the number of international conference increases. Many students study simultaneous interpretation due to the increased demand of simultaneous interpretation. Since simultaneous interpretation requires a lot of learning time, students majoring in translation perform the self learning. The paper based sight translation training system is a representative self learning method, but backtracking decreases the efficiency of self learning and it requires the help of the partner. To improve the learning efficiency, computer based sight translation training system has been proposed. However, since students uses the computer based sight translation training system only in a fixed area due to low mobility of computer, the utilization of the system decreases. In this paper, smart device based sight translation training system has been proposed to increase the utilization of the proposed system. Since smart device has lower computing capabilities than the computer, we have proposed algorithms to deal with the low performance. We implement and evaluate the functionalities of the proposed training system.

• ETRI Journal
• /
• v.28 no.6
• /
• pp.822-825
• /
• 2006

Based on superimposed training methods, a novel time-varying multipath channel estimation scheme is proposed for orthogonal frequency division multiplexing systems. We first develop a linear least square channel estimator, and meanwhile find the optimal superimposed sequences with respect to the channel estimates' mean square error. Next, a low-rank approximated channel estimator is obtained by using the singular value decomposition. As demonstrated in simulations, the proposed scheme achieves not only better performance but also higher bandwidth efficiency than the conventional pilot-aided approach.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1994.04a
• /
• pp.587-595
• /
• 1994

The use of flight training simulators has been increased remarkably during the past 20 years stimulated by the benefit that simulators can contribute to safety and training efficiency. Simulators have incorporated higher and higher levels of reality to provide more realistic training. Now, however, users are beginning to question how such reality is enough. Especially, it has raised question about the need for a notion system in the aircraft simulator. To help answer this question, this report describes notion cue requirements and types of motion system, and finally presents the assessment of the effectiveness of three different notion system alternatives.

• Korean Management Science Review
• /
• v.14 no.2
• /
• pp.161-184
• /
• 1997

This research examines the importance of IT training/education, present situation and possible suggestion for the successful training/education. The research method adopts a comparative analytical approach based on questionnaire survey responses from three work groups - managers, employees, and union representatives - drawn from five sample Korean banks. The evidence indicates that all three groups agree that IT improves banking efficiency and reduces job repetitiveness, but their job satisfaction level with IT-based work is surprisingly very low. The main reasons are mainly lack of training/education and poor user manuals. Also the research shows that most respondents would like to get further training/education to more adequately fit them for their jobs. Those from banks which invested in continuing training/education revealed more positive work attitudes and higher job satisfaction.