• Journal of Intelligence and Information Systems
• /
• 제26권2호
• /
• pp.105-129
• /
• 2020

This study uses corporate data from 2012 to 2018 when K-IFRS was applied in earnest to predict default risks. The data used in the analysis totaled 10,545 rows, consisting of 160 columns including 38 in the statement of financial position, 26 in the statement of comprehensive income, 11 in the statement of cash flows, and 76 in the index of financial ratios. Unlike most previous prior studies used the default event as the basis for learning about default risk, this study calculated default risk using the market capitalization and stock price volatility of each company based on the Merton model. Through this, it was able to solve the problem of data imbalance due to the scarcity of default events, which had been pointed out as the limitation of the existing methodology, and the problem of reflecting the difference in default risk that exists within ordinary companies. Because learning was conducted only by using corporate information available to unlisted companies, default risks of unlisted companies without stock price information can be appropriately derived. Through this, it can provide stable default risk assessment services to unlisted companies that are difficult to determine proper default risk with traditional credit rating models such as small and medium-sized companies and startups. Although there has been an active study of predicting corporate default risks using machine learning recently, model bias issues exist because most studies are making predictions based on a single model. Stable and reliable valuation methodology is required for the calculation of default risk, given that the entity's default risk information is very widely utilized in the market and the sensitivity to the difference in default risk is high. Also, Strict standards are also required for methods of calculation. The credit rating method stipulated by the Financial Services Commission in the Financial Investment Regulations calls for the preparation of evaluation methods, including verification of the adequacy of evaluation methods, in consideration of past statistical data and experiences on credit ratings and changes in future market conditions. This study allowed the reduction of individual models' bias by utilizing stacking ensemble techniques that synthesize various machine learning models. This allows us to capture complex nonlinear relationships between default risk and various corporate information and maximize the advantages of machine learning-based default risk prediction models that take less time to calculate. To calculate forecasts by sub model to be used as input data for the Stacking Ensemble model, training data were divided into seven pieces, and sub-models were trained in a divided set to produce forecasts. To compare the predictive power of the Stacking Ensemble model, Random Forest, MLP, and CNN models were trained with full training data, then the predictive power of each model was verified on the test set. The analysis showed that the Stacking Ensemble model exceeded the predictive power of the Random Forest model, which had the best performance on a single model. Next, to check for statistically significant differences between the Stacking Ensemble model and the forecasts for each individual model, the Pair between the Stacking Ensemble model and each individual model was constructed. Because the results of the Shapiro-wilk normality test also showed that all Pair did not follow normality, Using the nonparametric method wilcoxon rank sum test, we checked whether the two model forecasts that make up the Pair showed statistically significant differences. The analysis showed that the forecasts of the Staging Ensemble model showed statistically significant differences from those of the MLP model and CNN model. In addition, this study can provide a methodology that allows existing credit rating agencies to apply machine learning-based bankruptcy risk prediction methodologies, given that traditional credit rating models can also be reflected as sub-models to calculate the final default probability. Also, the Stacking Ensemble techniques proposed in this study can help design to meet the requirements of the Financial Investment Business Regulations through the combination of various sub-models. We hope that this research will be used as a resource to increase practical use by overcoming and improving the limitations of existing machine learning-based models.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• 제18권3호
• /
• pp.224-238
• /
• 2008

The aim of this study is to confirm the physicochemical property and hazard of thinner (012), which is a diluent of enamel paint used for floor coating for waterproofing and oil painting for the outer wall. The literatures of physicochemical property and hazard of thinner were surveyed and its physicochemical property were evaluated. And then, the inhalation toxicity of thinner affecting the central nervous system and reproductive organs in rats were examined by subchronic (6 h./day. 5 days/ week for 13 weeks) inhalation test. 1) According to the 13-week subchronic inhalation test, there were no significant changes in clinical test and body weight. However, a significant evidence of toxicity was observed in the hematological test and organ weight such as heart, kidney, liver and brain (p<0.01) in the 200 ppm and 1,000 ppm exposure groups in a dose response manner. In the histopathology analysis, there were no significant evidence of toxicity. Therefore, thinner was not classified as an organ targeted toxic agent. In case of Harmfulness, it could be classified as a chronic toxic agent 3($500 ppm/4hr, rat). 2) The reproductive toxicity such as extension of the period of estrous cycle, reduction of serum estradiol concentration and increase of frequency of the abnormal sperm was observed in the 1,000 ppm exposed animals. 3) The result of the physicochemical property of the test material showed that the specific gravity was 0.793, boiling point $155.8^{\circ}C$, steam pressure 2.1 kPa, ignition point $34.5^{\circ}C$, and spontaneous ignition point $280^{\circ}C$. The endothermic and exothermic values were 371.4 J/g and 159.1 J/g. respectively. The explosion limit was 214 mg/l. These data showed that thinner could be classified as an explosion agent level 1.2 and ignitive liquid agent 3 ($23-60^{\circ}C$) according to the notification No. 2008-1 of the Labor Ministry, "Classifying Standard of Chemical Materials."

• Journal of Convergence for Information Technology
• /
• 제10권9호
• /
• pp.206-214
• /
• 2020

This study analyzed the impact of direct and indirect subsidies on profitability in general hospitals in Korea. The data were collected from medical institution accounting information disclosure system of 270 general hospitals from 2016 to 2018. The analysis index used the ratio of net profit to business revenue for profitability, and Subsidies index the ratio of subsidies to business revenue(state subsidies for facility investment purposes, Fund related to essential business, research revenue and contribution revenue). According to the study, the ratio of state subsidies, which are direct subsidies, was very high at 57.30 percent for public institution hospitals. ratio of Fund related to essential business, which is a tax reduction effect with indirect subsidies, had the highest at 6.69 percent for Private University Hospitals. which are Indirect subsidies for deficit or operational assistance, research revenue ratio had the highest 2.8 percent for National University Hospitals, contribution revenue ratio had the highest 36.4 percent for public institutions. As a result of looking at the impact of subsidies on profitability, Nation University Hospitals had the lower the ratio of Fund related to essential business and the ratio of research revenue, the higher the net profit ratio of Business revenue. Medical Corporation Hospitals and Foundation Hospitals had the higher the ratio of Fund related to essential business to business revenues, the higher the net profit ratio of business revenue. These results mean that profitability may fluctuate depending on the utilization of funds related to essential business.

• Journal of Digital Convergence
• /
• 제15권11호
• /
• pp.285-295
• /
• 2017

This study aims to provide basic data on useful functional ingredients in red ginseng by studying the anti-inflammatory and anti-cancer effects of convergence of ginsenoside Rh2(Rh2) and compound K(CK) isolated from amplified red ginseng. Therefore we examined cytotoxicity in Hep3B, activity of IL-6 induced STAT3 luciferase and survival concentration of cells in B16F10 and HaCa T. According to the experimental results, when the Rh2 and CK mixture were 10 ug/ml, there was no cytotoxicity in Hep3B cells and the anti-inflammatory effect of IL-6 reduction ratio was 102%. In addition, Rh2 and CK mixture were observed to be toxic in melanoma cell line B16F10 and HaCa T (human keratinocyte) at 50 uM. FACS(fluorescence activated cell sorting) analysis showed that annexin V was not expressed and melanoma cells and keratinocyte were desorbed and killed. It can be assumed that the mechanism of killing through this phenomenon is due to the cell death of anoikis-type, and it is necessary to study the changes of cell adhesion proteins in the future in order to clarify the cell death signal system.

• Journal of the Korea Concrete Institute
• /
• 제29권4호
• /
• pp.379-387
• /
• 2017

The objective of this study is to develop sustainable PVA fiber-reinforced cementitious composites (FRCCs) that could exhibit comparable strength level to normal PVA FRCCs with no recycled materials. To evaluate mechanical properties of the FRCCs, compressive, flexural and direct tensile tests were conducted. In addition to the test, to calculate amount of carbon dioxide ($CO_2$) emission at the stage of manufacturing the FRCCs, life cycle inventory data base (LCI DB) were referenced from domestic and Japan. From the test results, the mechanical properties such as compressive, flexural and direct tensile strengths were decreased as the replacement ratio of recycled materials increased. And it was determined that the amount of $CO_2$ emission was reduced for the specimens with higher water-binder ratio (W/B) and replacement ratios. It was also found that binder intensity ($B_i$) value was higher as replacement ratio of fly ash (FA) increased. This result means that larger amount of FA is need to deliver one unit of a given performance indicator (1 MPa of strength) of FRCCs compared to that of ordinary portland cement (OPC). As a result, it could be concluded that FRCCs with W/B 45% replaced by FA 25% and recycled sand (RS) 25% is desirable for both target performance and $CO_2$ emission.

• International Journal of High-Rise Buildings
• /
• 제6권3호
• /
• pp.227-235
• /
• 2017

This paper describes some of the challenges for structural design of a mid-story seismic isolated high-rise building, which is located near Tokyo station, completed in 2015. The building is a mixed-use complex and encompasses three volumes: one substructure including basement and lower floors, and a pair of seismic isolated superstructures on the substructure. One is a 136.5m high Main Tower (office use), and the other is a 98.5 m high South Tower (hotel use). The seismic isolation systems are arranged in the $3^{rd}$ floor of the Main Tower and $5^{th}$ floor of the South Tower, so that we call this isolation system as the mid-story seismic isolation. The primary goal of the structural design of this building was to secure high seismic safety against the largest earthquake expected in Tokyo. We adopted optimal seismic isolation equipment simulated by dynamic analysis to minimize building damage. On the other hand, wind-induced vibration of a seismic isolated high-rise building tends to be excited. To reduce the vibration, the following strategies were adopted respectively. In the Main Tower with a large wind receiving area, we adopted a mechanism that locks oil dampers at the isolation level during strong wind. In the South Tower, two tuned mass dampers (TMDs) are installed at the top of the building to control the vibration. In addition, our paper will also report the building performance evaluated for wind and seismic observation after completion of the building. In 2016, an earthquake of seismic intensity 3 (JMA scale) occurred twice in Tokyo. The acceleration reduction rate of the seismic isolation level due to these earthquakes was approximately 30 to 60%. These are also verified by dynamic analysis using observed acceleration data. Also, in April 2016, a strong wind exceeding the speed of 25m/s occurred in Tokyo. On the basis of the record at the strong wind, we confirmed that the locking mechanism of oil damper worked as designed.

• Journal of the Korea Concrete Institute
• /
• 제27권4호
• /
• pp.427-434
• /
• 2015

The objective of this study is to establish an rational mixture-proportioning procedure for low-$CO_2$ concrete using supplementary cementitious materials (SCMs) achieving the targeted $CO_2$ reduction ratio as well as the conventional requirements such as initial slump, air content, and 28-day compressive strength of concrete. To evaluate the effect of SCM level on the $CO_2$ emission and compressive strength of concrete, a total of 12537 data sets were compiled from the available literature and ready-mixed concrete plants. The amount of $CO_2$ emission of concrete was assessed under the system boundary from cradle to concrete production stage at a ready-mixed concrete plant. Based on regression analysis using the established database, simple equations were proposed to determine the mixture proportions of concrete such as the type and level of SCMs, water-to-binder ratio, and fine aggregate-to-total aggregate ratio. Furthermore, the $CO_2$ emissions for a given concrete mixture can be straightforwardly calculated using the proposed equations. Overall, the developed mixture-proportioning procedure is practically useful for determining the initial mixture proportions of low-$CO_2$ concrete in the ready-mixed concrete field.

• Journal of Technology Innovation
• /
• 제19권1호
• /
• pp.1-27
• /
• 2011

This paper aims to measure and analyze R&D productivities and efficiencies of 17 major OECD countries including Korea over the 1984-2008 period by using the Malmquist Productivity Index and Data Envelopment Analysis, classifying R&D performance into an output and outcome aspects. It also searches the Korea's current status and characteristics in each R&D stage to enhance Total Factor Productivity (TFP) compared with other developed countries. Our major findings are the followings: (i) Korea's productivity index of R&D input vis-a-vis R&D output is very high (13.39% annual growth rate) compared with those of major advanced countries, whereas the annual average of efficiency index is very low (0.33), i.e. Korea's technical efficiency index has risen to 0.83 at the last time series started at 0.10 point and come up to the level of major advanced countries. (ii) the Korea's productivity index of R&D output vis-a-vis R&D outcome is very low (14.02% annual reduction rate) compared with those of major advanced countries, whereas the annual average of efficiency index is very high (0.22), i.e. Korea's integrated frontier technical efficiency index has dropped to 0.057 at the last time series started at 1.00 point and coming up to the level of major advanced countries. (iii) The productivity of R&D input vis-a-vis R&D outcome is positively correlated with that of R&D output vis-a-vis R&D outcome and the growth of R&D input factors. In a nutshell, it implicates that the effort to take advantage of R&D outputs, namely establishing the diffusion and commercialization system of technical knowledge to the level of developed countries, should be strengthened over that on the growth of R&D investment and output for enhancing R&D productivity and efficiency in Korea.

• International Commerce and Information Review
• /
• 제13권3호
• /
• pp.77-95
• /
• 2011

This paper investigates the relationship between export and economic variables such as trade insurance, world economy activity, relative price, unemployment rate, exchange rate volatility, using monthly data. I employ Johansen cointegration methodology since the model must be stationary to avoid the spurious results. The results indicate that there is a long-run relationship between export and variables. Also, the empirical analysis of cointegrating vector using the CCR, DOLS, FMOLS reveals that the increases of trade insurance has positive relations and the increases of exchange rate volatility have negative relations with export. Especially, DOLS based on Monte Carlo simulations, of this estimator being superior in small samples compared to a number of alternative estimators, as well as being able not only to accommodate higher orders of integration but also to account for possible simultaneity within regressors of a potential system. This paper also applies impulse-response functions to get the additional information regarding the responses of the export to the shocks of the variables. The result indicates that export positively to trade insurance and then decay fast compare with exchange rate volatility. Consequently, trade insurance plays the role of trade policy for export promotion in Korea. Whereas, increase of exchange risk result in reduction of export. Therefore, the support of trade insurance should be expanded and the stabilization of the foreign exchange market must be done for the export promotion.

• Journal of Korean Society of Transportation
• /
• 제32권2호
• /
• pp.83-92
• /
• 2014

Weather condition is one of the crucial factors affecting travelers' mode choice. Nevertheless, there are numerous indefinite traffic phenomena under various weather conditions. This study was conducted to verify the hypothesis that transit riderships decrease as precipitation increases. To clarify the relationship between precipitation and transit ridership, a seemingly unrelated regression model was employed with data such as daily precipitation and daily transit riderships of 3 transit modes (bus, metro, and shuttle bus) collected in Busan for recent 24 months. The estimation results show that transit riderships decreased as the daily precipitation increased when the daily precipitation is greater or equal to 10mm/day (0.169%, 0.101%, and 0.172% reduction in bus, metro, and shuttle bus riderships, respectively, when the daily precipitation increased by 1mm). When comparing the impact of precipitation on transit riderships by modes using a cross-equation parameter restriction test, the decrease in metro ridership is relatively insensitive to the change in precipitation. However, the negative coefficient of precipitation in the metro ridership estimation model indicates that the transit users in Busan may alter their mode to taxi or automobile and/or may give up the trip itself in bad weather condition.