• Journal of Ginseng Research
• /
• v.8 no.2
• /
• pp.167-171
• /
• 1984

Soft X-ray is useful to identify the quality of fresh ginseng causing the inside cavity or white pan of red ginseng. The portion of low mass density identified by the difference in absorption of soft X-ray showed lower dry matter density and less or no response to iodine test indicating less accumulation or excess consumption of starch. The inside white part of red ginseng absorbed less X-ray than the normal part did. Probability for identification of the inside cavity or white at fresh ginseng was rather high (80-90%) in screen observation than f'3m reading and seemed to be increased further by using the developed screen and with training. The inside white of red ginseng appeared to be due to starch deficiency. Dry matter density appeared to be better than fresh weight density for the quality criterion.

• Proceedings of the KIEE Conference
• /
• 1996.11a
• /
• pp.447-449
• /
• 1996

This paper describes an algorithm for short term load forecasting using the Kohonen neural network. Single layer Kohonen neural network presents a lot of advantageous features for practical application. It takes less training time compared to other networks such as BP network, and moreover, its self organized feature can amend the distorted data. The originality of proposed approach is to use a Kohonen map toclassify data representing load patterns and to use directly the information stored in the weight vectors of the Kohonen map to pridict the load. Proposed method was tested with KEPCO hourly record(1993-1995) show better forecasting results compared with conventional exponential smoothing method.

• Journal of the Korea Safety Management & Science
• /
• v.22 no.2
• /
• pp.47-55
• /
• 2020

This study analyzed ignition probability about Lithium-polymer batteries of what variously were being produced wearable devices recently. The study analyzed ignition probability by PCM(Protection Circuit Module) operating state and overcharged, over-discharged, exposed to high temperatures of Lithium polymer batteries, analyzing wearable devices on the market. Then it classified experimental results to implement analysis comparison about weight, X-ray imaging, battery decomposition. With these experiments, the study analyzed combustion-possibility and fire patterns. These statistics will be used to measure and verify the cause of a fire when identify wearable devices using Lithium-polymer batteries.

• Corrosion Science and Technology
• /
• v.20 no.3
• /
• pp.143-151
• /
• 2021

In this study, optimal shot peening process conditions were investigated for improving the cavitation erosion resistance of gray cast iron under a marine environment. Shot peening was performed with variables of injection pressure and injection time. The durability was then evaluated through cavitation erosion test which was conducted according to the modified ASTM G-32 standard. The tendency of cavitation erosion damage according to shot peening process condition was investigated through weight loss rate, surface and cross-sectional analysis of the specimen before and after the test. As a result, the shot peening process condition that could minimize cavitation erosion was when the injection pressure was the lowest and when the injection time was the shortest. This was because the flake graphite exposed on the gray cast iron surface could be easily removed under such condition. Therefore, the notch effect can be prevented by surface modification. In addition, the cavitation erosion damage mechanism of gray cast iron was discussed in detail.

• East Asian Economic Review
• /
• v.25 no.4
• /
• pp.403-424
• /
• 2021

AdaBoost tweaks the sample weight for each training set used in the iterative process, however, it is demonstrated that it provides more correlated errors as the boosting iteration proceeds if models' accuracy is high enough. Therefore, in this study, we propose a novel way to improve the performance of the existing AdaBoost algorithm by employing heterogeneous models and a stochastic twist. By employing the heterogeneous ensemble, it ensures different models that have a different initial assumption about the data are used to improve on diversity. Also, by using a stochastic algorithm with a decaying convergence rate, the model is designed to balance out the trade-off between model prediction performance and model convergence. The result showed that the stochastic algorithm with decaying convergence rate's did have a improving effect and outperformed other existing boosting techniques.

• Electronics and Telecommunications Trends
• /
• v.37 no.1
• /
• pp.53-62
• /
• 2022

With increasing size of transformer-based neural networks, a light-weight algorithm and efficient AI accelerator has been developed to train these huge networks in practical design time. In this article, we present a survey of state-of-the-art research on the low-precision computational algorithms especially for floating-point formats and their hardware accelerator. We describe the trends by focusing on the work of two leading research groups-IBM and Seoul National University-which have deep knowledge in both AI algorithm and hardware architecture. For the low-precision algorithm, we summarize two efficient floating-point formats (hybrid FP8 and radix-4 FP4) with accuracy-preserving algorithms for training on the main research stream. Moreover, we describe the AI processor architecture supporting the low-bit mixed precision computing unit including the integer engine.

• Steel and Composite Structures
• /
• v.46 no.1
• /
• pp.107-114
• /
• 2023

A new intelligent adaptive control scheme was proposed that combines observer disturbance-based adaptive control and fuzzy adaptive control for a composite structure with a mass-adjustable damper. The most important advantage is that the control structures do not need to know the uncertainty limits and the interference effect is eliminated. Three adjustable parameters in LMI are used to control the gain of the 2D fuzzy control. Binary performance indices with weighted matrices are constructed to separately evaluate validation and training performance using the revalidation learning function. Determining the appropriate weight matrix balances control and learning efficiency and prevents large gains in control. It is proved that the stability of the control system can be ensured by a linear matrix theory of equality based on Lyapunov's theory. Simulation results show that the multilevel simulation approach combines accuracy with high computational efficiency. The M-TMD system, by slightly reducing critical joint load amplitudes, can significantly improve the overall response of an uncontrolled structure.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.24-26
• /
• 2022

시스톨릭 어레이는 DNN training 등 인공지능 연산의 대부분을 차지하는 행렬 곱셈을 수행하기 위한 하드웨어 구조로 많이 사용되지만, sparsity 가 높은 행렬을 연산할 때 불필요한 동작으로 인해 효율성이 크게 떨어진다. 본 논문에서 제안된 유동적 시스톨릭 어레이는 matrix condensing, weight switching, 그리고 direct output path 의 방법과 구조를 통해 sparsity 가 높은 행렬 곱셈의 수행 사이클을 줄일 수 있다. 시뮬레이션을 통해 기존 시스톨릭 어레이와 유동적 시스톨릭 어레이의 성능을 비교하였으며 8×8, 16×16, 32×32 의 크기를 가진 행렬을 동일 크기의 시스톨릭 어레이로 연산하였을 때 필요 사이클 수를 최대 12 사이클 절감할 수 있는 것을 확인하였다.

• Journal of Life Science
• /
• v.20 no.9
• /
• pp.1394-1401
• /
• 2010

The purpose of this study was to examine the effect of 12 wk of aerobic exercise on lipid profiles, antioxidant enzyme activities, oxidative products, and autonomic nervous activity (ANA) in children with obesity. We studied 16 children with obesity and 19 age-matched normal weight controls over a period of 12 wk, during which time moderate intense running exercise was performed. Measurements included peak oxygen uptake, body composition, blood lipid profiles, ox-LDL, 8-OHdG, SOD, GPx activities, total mRNA, and ANA. There were no differences in body weight between periods in the OW group, but body weight increased after 12 wk in OR and CO groups. There were no differences in WHR between periods in the OR and CO groups, but the WHR values decreased after 12 weeks in the OW group. In the obese group, the baseline TG was higher than in the controls (p<0.05), while the ANA level was lower. There were differences in antioxidant enzyme gene expressions between periods in all groups. In conclusion, oxidative damage and antioxidant enzyme activities in obese children were found to be similar to those of normal weight children. However, TG was higher and ANA was lower in obese children than in normal weight children. These results indicated that increased TG and decreased ANA levels begins in childhood in obese patients. Also, regular aerobic exercise may modify the antioxidant enzyme gene expression in early life.

• Journal of Navigation and Port Research
• /
• v.39 no.1
• /
• pp.77-82
• /
• 2015

Slamming means that the hull hits the waves and receives impact pressure. This slamming effect may cause harm to people and when you put the hull at risk. so it is very harmful for cargo safety. Therefor slamming impact pressure should be fully considered in ship designing. In this study the model of wedged type structure are produced aimed to simulate a free fall that the experiments were carried out on different weight and free fall height. The flow field has been obtained by 2-frame grey level cross correlation PIV(Particle Image Velocimetry) method and experiment was divided into water entry and water exit. The impact pressure of free fall structure by a pressure acquisition system apply to dewetron system. The angles between a model and the water surface are adapted $15^{\circ}$ respectively. The weight change of models was given as 1.5, 1.8 and 2.0kg. To study slamming phenomenon for free fall height the experiments were carried out by the free fall height of 100, 200 and 300mm. The experimental value of the impact pressure according to the changes in weight was increase impact pressure in proportion to the increase in weight and higher free falling height has also influenced the increase in impact pressures.