• Journal of the Korean Society of Marine Environment & Safety
• /
• v.24 no.5
• /
• pp.511-518
• /
• 2018

Research indicates, about 80% of maritime accidents are caused by human error. Further investigation of the human factors behind maritime casualties is essential in order to establish preventive measures. The main purpose of this study is to identify and analyze human factors behind maritime traffic-related accidents using the m-SHEL model. Since the m-SHEL model used in other fields is based on generic human factors, it has expanded in this study to accommodate ship operating systems and define human factors. In addition, the validity of the expanded model was verified by reliability analysis using SPSSWIN. A classified table for this extended m-SHEL model was then used to analyze human factors behind maritime traffic-related accidents extracted from a written verdict by the Korean Maritime Safety Tribunal. Human factors were arranged in the order L, L-E, L-H, L-m, L-L, and L-S. This paper contributes to the prevention of maritime traffic-related accidents caused by human factors by presenting useful analytical results that can be applied to build a maritime safety management system.

• Journal of Bio-Environment Control
• /
• v.25 no.3
• /
• pp.146-152
• /
• 2016

Due to complicate canopy structures of Irwin mangoes grown in greenhouses, it is difficult to determine their growth parameters accurately. Leaf area, leaf fresh weight, and leaf dry weight are widely used as indicators to diagnose the tree growth. Therefore, it is necessary to establish models that can non-destructively estimate these growth indicators. The objective of this study was to establish regression models to estimate leaf area, leaf fresh weight, and leaf dry weight of Irwin mangoes (Mangifera indica L. cv. Irwin) by using leaf length, leaf width, petiole length, and SPAD value. The input values of leaf length, leaf width, petiole length, and SPAD value of 6-year old Irwin mangoes were measured, and the corresponding output values of leaf area, leaf fresh weight, and leaf dry weight were also measured. After 14 models were selected among the existing models, coefficients of the models were estimated by regression analysis. Three models with higher $R^2$ and lower RMSE values selected. In validation the $R^2$ values for the selected models were 0.967, 0.743, and 0.567 in the leaf area, leaf fresh weight, and leaf dry weight models, respectively. It is concluded that this models will be helpful to conveniently diagnose the growth of the Irwin mango.

• The Journal of the Convergence on Culture Technology
• /
• v.8 no.6
• /
• pp.569-580
• /
• 2022

The Minister of Education has officially announced that the Ministry of Education will promote the integration of early childhood education and care. The core of integration is the integration of kindergartens and child care centers, and it is necessary to develop the specific model, so this study was conducted. After the discussion of integration, the discussion has focused on the "Young Children School", and there were differences of opinion on specific issues. Therefore, this study was conducted to propose the specific model for the "Young Children School" after kindergartens and child care centers were integrated with the Ministry of Education. The research question is to find out what the model of the 0-5-year-old young children school model centered on the Ministry of Education. In this study, assuming that integration of early childhood education and care was realized, a plan for integrating early childhood education institutions by type of establishment was proposed. As a conclusion of this study, first, a model of a 0-5-year-old young children school centered the Ministry of Education under the government responsibility is required. Second, a detailed school model that reflects the characteristics of each type of establishment should be developed. Finally, specific measures such as the reform of the teacher training system, the after-school course operation model, and the integration plan with the office of education and local governments should be presented during the young children school system.

• Journal of the Korea Society of Computer and Information
• /
• v.29 no.7
• /
• pp.11-20
• /
• 2024

In this paper, we propose the BAG (Bit-width Aware Generator) and the Intermediate Layer Knowledge Distillation using Channel-wise Attention to reduce the knowledge gap between a quantized network, a full-precision network, and a generator in GDFQ (Generative Data-Free Quantization). Since the generator in GDFQ is only trained by the feedback from the full-precision network, the gap resulting in decreased capability due to low bit-width of the quantized network has no effect on training the generator. To alleviate this problem, BAG is quantized with same bit-width of the quantized network, and it can generate synthetic images, which are effectively used for training the quantized network. Typically, the knowledge gap between the quantized network and the full-precision network is also important. To resolve this, we compute channel-wise attention of outputs of convolutional layers, and minimize the loss function as the distance of them. As the result, the quantized network can learn which channels to focus on more from mimicking the full-precision network. To prove the efficiency of proposed methods, we quantize the network trained on CIFAR-100 with 3 bit-width weights and activations, and train it and the generator with our method. As the result, we achieve 56.14% Top-1 Accuracy and increase 3.4% higher accuracy compared to our baseline AdaDFQ.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.42 no.8
• /
• pp.692-700
• /
• 2014

In this work, satellite FE (finite element) model updating for the prediction of the effect of micro-vibration is described. In the case of satellites launched in low earth orbit, high agility and more mission accomplishments are required by the customer in order to procure many images from satellites. To achieve the goal, many mechanisms, including high capacity wheels and antennas with multi-axis gimbals have been widely adopted, but they become a source of micro-vibration which could significantly deteriorate the quality of images. To investigate the effect due to the micro-vibration in orbit on the ground, a prediction is conducted through an integrated model coupling the measured jitter sources with FE (finite element) model. Before prediction, the FE model is updated to match simulation results with the modal survey test. Subsequently, the quality of FE model is evaluated in terms of frequency deviation error, the resemblance of mode shapes and FRFs (frequency response functions) between test and analysis.

• Journal of Environmental Impact Assessment
• /
• v.27 no.5
• /
• pp.509-520
• /
• 2018

The impact of meteorological condition on surface $PM_{10}$ concentrations in South Korea was quantitatively simulated from 2010 to 2014 using WRF (ver.3.8.1) and CMAQ (5.0.2) model. The result showed that seasonal standard deviations of PM10 induced by change of weather conditions were $4.8{\mu}g/m^3$, $1.7{\mu}g/m^3$, $1.7{\mu}g/m^3$, $4.2{\mu}g/m^3$ for spring, summer, autumn and winter compared to 2010, respectively, with the annual mean standard deviation of about $2.6{\mu}g/m^3$. The results of 18 regions in South Korea showed standard deviation of more than $1{\mu}g/m^3$ in all regions and more than $2{\mu}g/m^3$ in Seoul, Northern Gyeonggi, Southern Southern Gyeonggi, Western Gangwon and Northern Chungcheong in South Korea.

• The Journal of the Korea Contents Association
• /
• v.13 no.2
• /
• pp.331-340
• /
• 2013

Recently there is a heated debate going on regarding the patient-doctor communication in the medical schools and medical service sector. Patient-doctor communication is an interactive communication made during the consultation session which is known to bring positive effect to both the patient and the doctor. Through this research, a doctor coaching model was developed by combining a method that would help the patient and doctor communicate better by increasing the doctor's communication skill and a coaching mechanism. Through the research, the doctor coaching model consists of 5 levels. First is the 'relationship creation' which would cause the doctor's interest and expectations toward coaching mechanism. Second is 'recognition of change' and this would cause to understand the problem and pros of the doctor's communication with the patient and set a direction regarding the coaching. Third is 'understanding the perspective' and this would lead the doctor to think from the patient's perspective. Fourth is 'increasing problem solving and communication skills' and this would set specific terms as to how the doctor can improve his communication skills. Fifth is 'goal setting and support' where goal regarding the improvements can be set and agreement regarding the ways to maintain and strengthen the advantage can be made. The developed doctor coaching model is most meaningful in a way that it has first adapted a coaching mechanism to improve patient-doctor communication. Also in cases where such will be utilized in the future medical service sector, it is expected to affect greatly the doctor's communication skill and patient sympathizing skills. Hereby it will contribute in increasing the patient's treatment satisfaction.

• Tunnel and Underground Space
• /
• v.30 no.3
• /
• pp.214-225
• /
• 2020

Uniaxial compressive strength (UCS) of rock is one of the important factors to determine the advance speed during shield TBM tunnel excavation. UCS can be obtained through the Geotechnical Data Report (GDR), and it is difficult to measure UCS for all tunneling alignment. Therefore, the purpose of this study is to predict UCS by utilizing TBM machine driving data and machine learning technique. Several machine learning techniques were compared to predict UCS, and it was confirmed the stacking model has the most successful prediction performance. TBM machine data and UCS used in the analysis were obtained from the excavation of rock strata with slurry shield TBMs. The data were divided into 8:2 for training and test and pre-processed including feature selection, scaling, and outlier removal. After completing the hyper-parameter tuning, the stacking model was evaluated with the root-mean-square error (RMSE) and the determination coefficient (R²), and it was found to be 5.556 and 0.943, respectively. Based on the results, the sacking models are considered useful in predicting rock strength with TBM excavation data.

• The Korean Journal of Pesticide Science
• /
• v.9 no.4
• /
• pp.279-286
• /
• 2005

The herbicidal activities against pre-emergence barnyard grass (Echinochloa crus-galli) by a series of new 2-(4-(6-chloro-2-benzoxazolyloxy)phenoxy)-N-phenylpopionamide derivatives as substrate molecule were studied using molecular holographic (H) quantitative structure activity relationships (HQSAR) methodology. From the based on the findings, the higher herbicidal active compounds are predicted by the derived HQSAR model. The best HQSAR model (VI-1) was derived from fragment distinction combination of atoms/bonds in fragment size, $7{\sim}10$bin. The herbicidal activities from atomic contribution maps showed that the activity will be able to increased according to the R-substituents variation of the N-phenyl ring and change of 6-chloro-2-benzoxazolyloxy group. Based on the results, the statistical results of the best HQSAR model (VI-1) exhibited the best pedictability and fitness for the herbicidal activities based on the cross-validated value ($q^2=0.646$) and non cross-validated value ($r^2_{ncv.}=0.917$), respectively. From the graphical analyses of atomic contribution maps, it was revealed that the lowest herbicidal activitics depends upon the 4-(6-chloro-2-benzoxazolyloxy)phenoxy group ($pred.pI_{50}=-3.20$). Particularly, the R=4-fluoro, X=isobutoxy substituent (P2) of (X)-phenoxy-N-(R)-phenylpropionamide derivative is predicted as the highest active compound ($pred.pI_{50}=9.12$).

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.6
• /
• pp.1099-1110
• /
• 2023

The recognition system of autonomous driving and robot navigation performs vision work such as object recognition, tracking, and lane detection after multi-sensor fusion to improve performance. Currently, research on a deep learning model based on the fusion of a camera and a lidar sensor is being actively conducted. However, deep learning models are vulnerable to adversarial attacks through modulation of input data. Attacks on the existing multi-sensor-based autonomous driving recognition system are focused on inducing obstacle detection by lowering the confidence score of the object recognition model.However, there is a limitation that an attack is possible only in the target model. In the case of attacks on the sensor fusion stage, errors in vision work after fusion can be cascaded, and this risk needs to be considered. In addition, an attack on LIDAR's point cloud data, which is difficult to judge visually, makes it difficult to determine whether it is an attack. In this study, image scaling-based camera-lidar We propose an attack method that reduces the accuracy of LCCNet, a fusion model (camera-LiDAR calibration model). The proposed method is to perform a scaling attack on the point of the input lidar. As a result of conducting an attack performance experiment by size with a scaling algorithm, an average of more than 77% of fusion errors were caused.