• Journal of Intelligence and Information Systems
• /
• v.27 no.1
• /
• pp.83-102
• /
• 2021

The government recently announced various policies for developing big-data and artificial intelligence fields to provide a great opportunity to the public with respect to disclosure of high-quality data within public institutions. KSURE(Korea Trade Insurance Corporation) is a major public institution for financial policy in Korea, and thus the company is strongly committed to backing export companies with various systems. Nevertheless, there are still fewer cases of realized business model based on big-data analyses. In this situation, this paper aims to develop a new business model which can be applied to an ex-ante prediction for the likelihood of the insurance accident of credit guarantee. We utilize internal data from KSURE which supports export companies in Korea and apply machine learning models. Then, we conduct performance comparison among the predictive models including Logistic Regression, Random Forest, XGBoost, LightGBM, and DNN(Deep Neural Network). For decades, many researchers have tried to find better models which can help to predict bankruptcy since the ex-ante prediction is crucial for corporate managers, investors, creditors, and other stakeholders. The development of the prediction for financial distress or bankruptcy was originated from Smith(1930), Fitzpatrick(1932), or Merwin(1942). One of the most famous models is the Altman's Z-score model(Altman, 1968) which was based on the multiple discriminant analysis. This model is widely used in both research and practice by this time. The author suggests the score model that utilizes five key financial ratios to predict the probability of bankruptcy in the next two years. Ohlson(1980) introduces logit model to complement some limitations of previous models. Furthermore, Elmer and Borowski(1988) develop and examine a rule-based, automated system which conducts the financial analysis of savings and loans. Since the 1980s, researchers in Korea have started to examine analyses on the prediction of financial distress or bankruptcy. Kim(1987) analyzes financial ratios and develops the prediction model. Also, Han et al.(1995, 1996, 1997, 2003, 2005, 2006) construct the prediction model using various techniques including artificial neural network. Yang(1996) introduces multiple discriminant analysis and logit model. Besides, Kim and Kim(2001) utilize artificial neural network techniques for ex-ante prediction of insolvent enterprises. After that, many scholars have been trying to predict financial distress or bankruptcy more precisely based on diverse models such as Random Forest or SVM. One major distinction of our research from the previous research is that we focus on examining the predicted probability of default for each sample case, not only on investigating the classification accuracy of each model for the entire sample. Most predictive models in this paper show that the level of the accuracy of classification is about 70% based on the entire sample. To be specific, LightGBM model shows the highest accuracy of 71.1% and Logit model indicates the lowest accuracy of 69%. However, we confirm that there are open to multiple interpretations. In the context of the business, we have to put more emphasis on efforts to minimize type 2 error which causes more harmful operating losses for the guaranty company. Thus, we also compare the classification accuracy by splitting predicted probability of the default into ten equal intervals. When we examine the classification accuracy for each interval, Logit model has the highest accuracy of 100% for 0~10% of the predicted probability of the default, however, Logit model has a relatively lower accuracy of 61.5% for 90~100% of the predicted probability of the default. On the other hand, Random Forest, XGBoost, LightGBM, and DNN indicate more desirable results since they indicate a higher level of accuracy for both 0~10% and 90~100% of the predicted probability of the default but have a lower level of accuracy around 50% of the predicted probability of the default. When it comes to the distribution of samples for each predicted probability of the default, both LightGBM and XGBoost models have a relatively large number of samples for both 0~10% and 90~100% of the predicted probability of the default. Although Random Forest model has an advantage with regard to the perspective of classification accuracy with small number of cases, LightGBM or XGBoost could become a more desirable model since they classify large number of cases into the two extreme intervals of the predicted probability of the default, even allowing for their relatively low classification accuracy. Considering the importance of type 2 error and total prediction accuracy, XGBoost and DNN show superior performance. Next, Random Forest and LightGBM show good results, but logistic regression shows the worst performance. However, each predictive model has a comparative advantage in terms of various evaluation standards. For instance, Random Forest model shows almost 100% accuracy for samples which are expected to have a high level of the probability of default. Collectively, we can construct more comprehensive ensemble models which contain multiple classification machine learning models and conduct majority voting for maximizing its overall performance.

• Journal of Food Hygiene and Safety
• /
• v.30 no.4
• /
• pp.343-349
• /
• 2015

A study survey about the rice cake producers completing the food hygiene education in Korea was investigated by characteristics of the rice cake business. The difference between their online and offline awareness of food hygiene education were compared. The average age of rice cake producers is 50 (40.1%), with a high school education (52.6%), 10-20 years of service (34.3%) showed the highest percentage. In relation to sales and work area, workshop personnel are engaged in two (79.5%), An area of less than $99.17m^2$ (92.0%), rent (60.2%) with most paying a monthly rental amount of less than 1 million won (54.8%). There were 228 accident cases in three years (an annual average of 2.4%), manufacturing, Processing the item number was less than 20 types of analysis (86.7%). Case of food hygiene education graduates are women, the lower the age, the higher the education level, was preferred online. Online education was chosen because of 'time, economic, convenience'(73.7%). Online graduates have recognized that health education is more conducive to business. There was no significant difference between the sales online and offline graduates. For hygienic management response was that online graduates are well above the 7.4% offline graduates. Online and offline graduates 60.7% appeared to be more satisfied than the previous training institutions.

• Journal of Korean Academy of Nursing
• /
• v.12 no.2
• /
• pp.80-90
• /
• 1982

This study was undertaken to determine the influence of a chronic disease on body image, and to show that body image is directly related to one's self concept. Body image is the concept of one's own body based on present and past perception, and is elated to one's self concept. Body image is a dynamic concept constantly changing throughout the life cycle but it changes greatly in illness, surgery, and accident. The child with a chronic disease experiences pain and immobilization due to illness and he/she experiences a strange environment in the hospital. illness often brings feeling of frustration and loss of self-esteem. Therefore this study was done to compare the body image of a child with a chronic disease（Asthma), with that of a normal healthy child, and to determine the relationship between the body image and self concept. The subjects in this study were 36 children being treated for asthma at the allergy clinic of Y University Hospital in Seoul (patient group) and 44 children attending elementary school in Kwanak Ku Seoul (normal healthy group). For the measurement of the body image, the researcher used Secord & Jourard's Body Cathexis Scale, and another scale which was constructed after reading about Osgood's Semantic Differential Method. For the measurement of the self concept, the researcher used Jacox & Stewart's Health Self Concept. The period for data collection was from October 7th to October 27th, 1982. The analysis of data was done by use of Percentage, t-test, Chi-square test, Pearson Correlation Coefficient and ANOVA, The results of the study were as follows: 1. The first hypothesis,“That the chronically ill (Asthma) child will have a more negative body image than the normal healthy child.”was supported. 2. The second hypothesis,“The more negative the body image, the lower the self concept.”was also supported. 3. The researcher failed to obtain statistically significant results in the analysis of the general characteristics which affect the body image except in the case of the older child as compare to the younger Child having a mole positive body image (r=.2751, r=.2481, p<.05). However it was found that, 1) Boy's have a more positive body image than girls (Mean=〔37.81, 141.09〕,〔37.00, 126.54〕), 2) The child who has been hospitalized has a more negative body image than the child who has never been hospitalized (Mean=〔33.25, 122.45〕,〔35.68, 129.93〕). 3) The younger the child when the disease is discovered and diagnosed, the more negative the body image (Onset of illness: Mean=〔31.44, 117.33〕,〔34.00, 103.50〕, 〔35.75, 140.38〕,〔36.33, 130.00〕, Time of Diagnosis: Mean=〔29.00, 117.33〕,〔33.89, 115.00〕,〔33.36, 124.93〕,〔37.10, 139. 20〕). In conclusion the chronically ill(.Asthma) child has a more negative body image than the normal healthy child, and the more negative the body image the lower the self concept. Therefore the concept of body image is useful in understanding the influences of chronic disease on body' image and self concept.

• Journal of Radiation Protection and Research
• /
• v.24 no.2
• /
• pp.65-72
• /
• 1999

A methodology for determining the optimal durations of the use of contaminated crops as feedstuffs of cattle was designed based on the cost-benefit analysis method. The results of application for pigs, an omnivorous cattle, were discussed for the hypothetical deposition of radionuclides on August 15 when a number of crops are fully developed in Korean agricultural conditions. For investigating the relative cost-effectiveness of the use of contaminated crops as feedstuffs, the net benefit was compared with the case of the direct disposal of contaminated crops. The time-dependent radionuclide concentration in crops after the deposition was predicted using a dynamic food chain model DYNACON. The net benefit from the actions was quantitatively evaluated in terms of cost equivalent of doses and monetary costs of implementing the action. It depended on a number of factors such as radionuclides, variety of crops supplied as feedstuffs and duration of the actions. The use of contaminated crops as feedstuffs was more cost effective for $^{90}Sr\;or\;^{131}I$ deposition than for $^{137}Cs$ deposition.

• Korean Security Journal
• /
• no.17
• /
• pp.33-51
• /
• 2008

According to dictionary, the meaning of protection is "guard and protect" that means protecting the Protectee's safety in case of sudden attack or various accident and Security means all protecting activity including Protectee and place where he is in or will be as comprehensively meaning of safe. As you see in the definition, Protection and security is the act to protect or will to protect from a security-threat act. A security-threat act can be discussed in the range of the concept of a criminal act in Criminal Law. A security-threat act is based on criminal act in Criminal Law, we are going to review such a security-threat act in a point of view in a sphere of learning in today's remarkable a brain-neuro science and cognitive science based on cognitive psychology, and to use an analysis on such a security-threat act to make a foundation for a establishment of the scientific preventive social security countermeasure. To do so, First of all we are going to review a security-threat act based on criminal act in Criminal Law in a point of protection police logic view. Next, we are going to introduce how cognitive science understand about act of man before we analyse a threat act as one of an act of man in cognitive science point of view. Finally, we are going to discuss the need of cognitive scientific analyse in order to establish the Scientific Preventive Social-security Countermeasure at the same time we are going to analyse a threat act in a cognitive scientific view.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.22 no.3
• /
• pp.293-310
• /
• 2020

The purpose of this study is to present the most effective smoke exhaust mode by comparing the quantitatively evaluated risks according to the smoke exhaust mode when a train fire occurs in a subway platform. Therefore, applying the typical subway platform as a model, train fire scenarios are developed with the evacuation start time and location of the fire train for each exhaust mode. The fire accident rates (F) are calculated and the number of fatalities (N) was quantitatively estimated by fire analysis and evacuation analysis for each scenario. In addition, the F/N curve compared with the social risk assessment criteria and the following conclusions were obtained. In the event of a train fire at the subway station platform, the evacuation must start up within 600 s in maximum to ensure the evacuees' safety. To secure evacuation safety, it is advantageous to operate the HVAC system of the platform in the air-supply mode at station without TVF. Comparing the F/N curve for each exhaust mode with the social risk criteria, it turned out that the risk significantly exceeds the social risk criteria in case of no mechanical ventilation. As a result, this paper shows that the ventilation mode in which TVF are exhausted and HVAC system is operated in the pressurized mode are the most effective smoke exhaust mode for ensuring evacuation safety.

• Fire Science and Engineering
• /
• v.32 no.6
• /
• pp.1-6
• /
• 2018

This study is to analyze the characteristics of the yellow insulation ring type of the CSST used for tubing when it is artificially deteriorated and damaged by burning. The CSST for tubing consists of a tube, protective coating, nut, yellow insulation ring, packing, and socket. In addition, it is thought that a yellow insulation ring and rubber packing were used to connect the tube and socket in order to improve the airtightness and insulation performance. The result of the verification of the data acquired from the tests in the 95% confidence interval shows that the Anderson-Darling (AD) and P value were analyzed to be 0.945 and 0.015, respectively. This confirms that the test data of the CSST for tubing is reliable. The analysis of the arithmetic mean of the insulation resistance of a CSST showed that the CSST damaged by burning by a torch, and the one damaged by electrical burning, was $16.7k{\Omega}$ (the greatest relatively) and $208{\Omega}$ (the lowest), respectively, while it was $1.72k{\Omega}$ in the case of a normal product. Therefore, the analysis result of the insulation resistance of the CSST collected from the scene of a fire can be utilized to examine the cause of damage by burning. In addition, it was found that when the maximum current of 97 A was applied to the CSST for about 5 s using a Primary Current Injection Test System (PCITS) the protective film and insulation ring of the CSST has no difference from that of a normal product. However, a part of the metal tube was melted.

• The Journal of the Korea Contents Association
• /
• v.19 no.9
• /
• pp.92-100
• /
• 2019

The purpose of this study is to analyze the 2019 Revised Nuri Curriculum and the 2015 Revised Curriculum for the comparative analysis of disaster safety education contents in the national level curriculum. The main results of this study are as follows: First, the content of disaster safety education for preschoolers in the 2019 Revised Nuri Curriculum was clarified as 'Knowing how to react in case of accident, fire, disaster, abuse, kidnap etc.' in 'Living safety' category of physical exercise and health sections. Second, in the 2015 Revised Curriculum, disaster safety education contents for elementary school students were found to have the difference between school year in same elementary school curriculum. For first and second graders, 'Disaster safety' was cleary presented in one of major categories of 'Living safety', which was newly created at the creative activity in the Revised 2015 Curriculum. On the other hand, for third to sixth graders, safety education is dispersed in the related subjects such as physical education, science, practical courses or in the creativity activity. Third, the safety contents in the Revised 2015 Curriculum for middle school and high school were also found to be dispersed in social studies, physical education, science, technology-home economics and creativity activity. The results of this study showed that the contents of disaster safety education in the national curriculum did not have continuity and systemicity from kindergarten to high school. Therefore, efforts should be made to ensure continuity and systemicity of disaster safety education contents when the national curriculum is revised in the future.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.7
• /
• pp.1111-1119
• /
• 2022

Tourism projects through islands in the waters of Sinan-gun became active, and as a result, a total of 13 marine bridges connecting islands were completed. However, the marine bridge constructed in the fairway is dangerous for traffic. Particularly, in the case of the marine bridge connecting two islands, the width of the fairway is extremely narrow, therefore the risk is higher. In this study, we evaluated the risk of collision between marine bridge piers and ships using the IALA Waterway Risk Assessment Program (IWRAP), a risk assessment model for port waterways, based on a maritime traffic survey on the coastal bridge in Sinan-gun. The results, indicated that No.1 Sinan bridge had the highest probability of collision and most of the transit ships were coastal passenger ships. In addition, No.1 Sinan bridge was the place where the most collision accidents occurred among the marine bridge piers in the target sea, and the cause this study was analyzed. An analysis of the satellite images of the sea environment of No.1 Sinan bridge using an image processing method, confirmed that obstacles that could not be seen in the chart existed nearby the bridge. As a result, traffic was observed to be concentrated in one direction, unlike two-way traffic, which is a method of inducing traffic of bridges to avoid obstacles. The risk cause analysis method using the image processing technique of this study is expected to be used as a basic research method for analyzing the risk factors of island bridge in the future.

• Composites Research
• /
• v.36 no.4
• /
• pp.253-258
• /
• 2023

Although the development of transportation means has realized the right to mobility for the disabled who have difficulty in moving, it can be said that the improvement of the safety of passengers with disabilities that can occur in a car accident is lower than that of ordinary passenger seats. In particular, in the case of a rear-end collision that can occur suddenly, it is a reality that disabled passengers are vulnerable to head and neck injuries. Therefore, in this study, a multi-layer headrest foam that divides the headrest into three parts in the coronal plane was proposed to improve the head and neck injury index of disabled passengers in the vehicle in the event of a rear-end collision of a wheelchair transport vehicle. A range of stress scale factors was selected to give various compressive characteristics of the foam through low-speed rear-end collision analysis through a simple model, and GA optimization was performed by specifying the range as a parameter. Through the optimization result, the phase relationship between HIC and NIC was analyzed according to the compression characteristics of the layers. HIC responded most sensitively to the compression characteristics of the front layer and NIC responded to the compression characteristics of the mid layer, and the compression characteristics of the rear layer showed the lowest. A normal headrest and an optimized multi-layer headrest were placed in the validation model to analyze the low-speed rear-end collision sled test, and HIC and NIC were derived lower in the multi-layer headrest than in the general headrest. The compression behavior of the multi-layer headrest was also clearly shown, and it was verified that the multi-layer headrest was effective in improving the injury index of the head and neck compared to the general headrest.