• Korean Journal of Remote Sensing
• /
• v.36 no.5_3
• /
• pp.1053-1066
• /
• 2020

Sulfur dioxide (SO₂) is primarily released through industrial, residential, and transportation activities, and creates secondary air pollutants through chemical reactions in the atmosphere. Long-term exposure to SO₂ can result in a negative effect on the human body causing respiratory or cardiovascular disease, which makes the effective and continuous monitoring of SO₂ crucial. In South Korea, SO₂ monitoring at ground stations has been performed, but this does not provide spatially continuous information of SO₂ concentrations. Thus, this research estimated spatially continuous ground-level SO₂ concentrations at 1 km resolution over South Korea through the synergistic use of satellite data and numerical models. A stacking ensemble approach, fusing multiple machine learning algorithms at two levels (i.e., base and meta), was adopted for ground-level SO₂ estimation using data from January 2015 to April 2019. Random forest and extreme gradient boosting were used as based models and multiple linear regression was adopted for the meta-model. The cross-validation results showed that the meta-model produced the improved performance by 25% compared to the base models, resulting in the correlation coefficient of 0.48 and root-mean-square-error of 0.0032 ppm. In addition, the temporal transferability of the approach was evaluated for one-year data which were not used in the model development. The spatial distribution of ground-level SO₂ concentrations based on the proposed model agreed with the general seasonality of SO₂ and the temporal patterns of emission sources.

• Journal of Intelligence and Information Systems
• /
• v.23 no.3
• /
• pp.139-153
• /
• 2017

Corporate bankruptcy can cause great losses not only to stakeholders but also to many related sectors in society. Through the economic crises, bankruptcy have increased and bankruptcy prediction models have become more and more important. Therefore, corporate bankruptcy has been regarded as one of the major topics of research in business management. Also, many studies in the industry are in progress and important. Previous studies attempted to utilize various methodologies to improve the bankruptcy prediction accuracy and to resolve the overfitting problem, such as Multivariate Discriminant Analysis (MDA), Generalized Linear Model (GLM). These methods are based on statistics. Recently, researchers have used machine learning methodologies such as Support Vector Machine (SVM), Artificial Neural Network (ANN). Furthermore, fuzzy theory and genetic algorithms were used. Because of this change, many of bankruptcy models are developed. Also, performance has been improved. In general, the company's financial and accounting information will change over time. Likewise, the market situation also changes, so there are many difficulties in predicting bankruptcy only with information at a certain point in time. However, even though traditional research has problems that don't take into account the time effect, dynamic model has not been studied much. When we ignore the time effect, we get the biased results. So the static model may not be suitable for predicting bankruptcy. Thus, using the dynamic model, there is a possibility that bankruptcy prediction model is improved. In this paper, we propose RNN (Recurrent Neural Network) which is one of the deep learning methodologies. The RNN learns time series data and the performance is known to be good. Prior to experiment, we selected non-financial firms listed on the KOSPI, KOSDAQ and KONEX markets from 2010 to 2016 for the estimation of the bankruptcy prediction model and the comparison of forecasting performance. In order to prevent a mistake of predicting bankruptcy by using the financial information already reflected in the deterioration of the financial condition of the company, the financial information was collected with a lag of two years, and the default period was defined from January to December of the year. Then we defined the bankruptcy. The bankruptcy we defined is the abolition of the listing due to sluggish earnings. We confirmed abolition of the list at KIND that is corporate stock information website. Then we selected variables at previous papers. The first set of variables are Z-score variables. These variables have become traditional variables in predicting bankruptcy. The second set of variables are dynamic variable set. Finally we selected 240 normal companies and 226 bankrupt companies at the first variable set. Likewise, we selected 229 normal companies and 226 bankrupt companies at the second variable set. We created a model that reflects dynamic changes in time-series financial data and by comparing the suggested model with the analysis of existing bankruptcy predictive models, we found that the suggested model could help to improve the accuracy of bankruptcy predictions. We used financial data in KIS Value (Financial database) and selected Multivariate Discriminant Analysis (MDA), Generalized Linear Model called logistic regression (GLM), Support Vector Machine (SVM), Artificial Neural Network (ANN) model as benchmark. The result of the experiment proved that RNN's performance was better than comparative model. The accuracy of RNN was high in both sets of variables and the Area Under the Curve (AUC) value was also high. Also when we saw the hit-ratio table, the ratio of RNNs that predicted a poor company to be bankrupt was higher than that of other comparative models. However the limitation of this paper is that an overfitting problem occurs during RNN learning. But we expect to be able to solve the overfitting problem by selecting more learning data and appropriate variables. From these result, it is expected that this research will contribute to the development of a bankruptcy prediction by proposing a new dynamic model.

• Earthquakes and Structures
• /
• v.5 no.2
• /
• pp.129-142
• /
• 2013

Linear and nonlinear time history analyses have been becoming more common in seismic analysis and design of structures with advances in computer technology and earthquake engineering. One of the most important issues for such analyses is the selection of appropriate acceleration time histories and matching these histories to a code design acceleration spectrum. In literature, there are three sources of acceleration time histories: artificial records, synthetic records obtained from seismological models and accelerograms recorded in real earthquakes. Because of the increase of the number of strong ground motion database, using and scaling real earthquake records for seismic analysis has been becoming one of the most popular research issues in earthquake engineering. In general, two methods are used for scaling actual earthquake records: scaling in time domain and frequency domain. The objective of this study is twofold: the first is to discuss and summarize basic methodologies and criteria for selecting and scaling ground motion time histories. The second is to analyze scaling results of time domain method according to ASCE 7-05 and Eurocode 8 (1998-1:2004) criteria. Differences between time domain method and frequency domain method are mentioned briefly. The time domain scaling procedure is utilized to scale the available real records obtained from near fault motions and far fault motions to match the proposed elastic design acceleration spectrum given in the Eurocode 8. Why the time domain method is preferred in this study is stated. The best fitted ground motion time histories are selected and these histories are analyzed according to Eurocode 8 (1998-1:2004) and ASCE 7-05 criteria. Also, characteristics of both near fault ground motions and far fault ground motions are presented by the help of figures. Hence, we can compare the effects of near fault ground motions on structures with far fault ground motions' effects.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1099-1107
• /
• 2017

In this study, we analyzed the determinants of wages of college graduates by using the data of "2014 Graduates Occupational Mobility Survey" conducted by Korea Employment Information Service. In general, wages contain two complex pieces of information about whether an individual is employed and the size of the wage. However, in many previous researches on wage determinants, sample selection bias tends to be generated by performing linear regression analysis using only information on wage size. We used the Heckman sample selection models for analysis to overcome this problem. The main results are summarized as follows. First, the validity of the Heckman's sample selection model is statistically significant. Male is significantly higher in both job probability and wage than female. As age increases and parents' income increases, both the probability of employment and the size of wages are higher. Finally, as the university satisfaction increases and the number of certifications acquired increased, both the probability of employment and the wage tends to increase.

• The Korean Journal of Community Living Science
• /
• v.15 no.3
• /
• pp.21-28
• /
• 2004

Marketing of traditional doenjang (a kind of fermented soybean product) is now focusing on its health advantages, being proven to have anti-viral, anti-cancer and anti-oxidant effects. The purpose of this study is to investigate the consumption patterns of traditional doenjang in the households managed by different generations of women living in Gyeonggi area. Six hundred housewives answered the questionnaire. Statistical analyses were performed on 590 subjects using SAS (ver 8.1). Chi-square tests and General Linear Models were used. The age distribution of housewives was as follows: 42.9% were in their 30s; 40.9% were in their 40s; and 16.2% were in their 50s. 57.5% of subjects graduated with high school education while 72.8% of subjects did housework only. Overall, 47.2%, prepared their doenjang themselves, while the remainder purchased it or received it from relatives. This percentage differed however according to age group, as self-preparation of doenjang was found in only 22% of housewives in their 30s, but increased to 83% of subjects in their 50s. 53.4% of subjects had their doenjang donated to them by relatives, compared with only 3.1 % of subjects in their 50s. Most of dishes using doenjang were soups and stews. Those housewives in their 50s made significantly higher use of doenjang in soup, wrapping vegetables (ssamjang), seasoning, and flat cake (jangttok). Doenjang dishes were prepared for the husband in the family 59.2% of the time, followed by housewives 15.4% of the time. Annual consumption of doenjang was 5.1kg. and Kanjang was 4.4L per household; there was no significant difference between the age groups respecting annual consumption. From such results, we may assume a decrease in home-made doenjang among younger generations and an increase in the amount of purchased doenjang. We can predict an increased need for doenjang of better quality. Also the development of new products such as ready-to-eat or fast food variants would better serve the trend towards convenience.

• Journal of Korean Institute of Industrial Engineers
• /
• v.6 no.2
• /
• pp.21-29
• /
• 1980

This paper deals with the capital budgeting problem of a firm where investments are risky and interrelated. The established models might be classified into two categories; One is the chance-constrained programming model and the other is the expected utility maximization model. The former has a rather limited objective function and does not consider the risk in direct manner. The latter, on the other hand, might lead to a wrong decision because it uses an approximate value of expected utility. This paper attempts to extend the applicability of the chance-constrained programming model by modifying its objective function into a more general form. The capital budgeting problem is formulated as a nonlinear 0-1 integer programming problem first, and is formulated into a linear 0-1 integer programming problem for finding a lower-bound solution of the original problem. The optimal solution of the original problem is then obtained by branch & bound algorithm.

• The Journal of Internal Korean Medicine
• /
• v.28 no.4
• /
• pp.872-885
• /
• 2007

Objective : Stroke is one of the most common causes of death in Korea. This study was done to evaluate the association of complete blood count (CBC) with the risk of hemorrhagic stroke and ischemic stroke. Methods : In 217-case patients with ischemic stroke or hemorrhagic stroke and 146 healthy control subjects without stroke, hypertension, diabetes mellitus, hyperlipidemia, or ischemic heart disease and 160 controls without ischemic stroke or hemorrhagic stroke, we tested and compared white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hgb), hematocrit (Hct) and platelet. These data were statically analyzed by general linear models and binary logistic regression analysis to get each adjusted odds ratio. Results :The level of WBC was significantly higher in all cases. The level of RBC, Hct and Hgb was significantly lower in patients of ischemic stroke. The level of platelet was significantly higher in patients of ischemic stroke. Conclusion : These results suggest high WBC may be a risk factor of hemorrhagic stroke and ischemic stroke and low RBC, low Hct, low Hgb and high platelet may be risk factors of ischemic stroke in Koreans.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.66 no.5
• /
• pp.833-842
• /
• 2017

In this study, the design methodology of radial basis function neural networks based on incremental K-means clustering is introduced for learning and processing the big data. If there is a lot of dataset to be trained, general clustering may not learn dataset due to the lack of memory capacity. However, the on-line processing of big data could be effectively realized through the parameters operation of recursive least square estimation as well as the sequential operation of incremental clustering algorithm. Radial basis function neural networks consist of condition part, conclusion part and aggregation part. In the condition part, incremental K-means clustering algorithms is used tweights of the conclusion part are given as linear function and parameters are calculated using recursive least squareo get the center points of data and find the fitness using gaussian function as the activation function. Connection s estimation. In the aggregation part, a final output is obtained by center of gravity method. Using machine learning data, performance index are shown and compared with other models. Also, the performance of the incremental K-means clustering based-RBFNNs is carried out by using PSO. This study demonstrates that the proposed model shows the superiority of algorithmic design from the viewpoint of on-line processing for big data.

• Asian-Australasian Journal of Animal Sciences
• /
• v.16 no.5
• /
• pp.685-692
• /
• 2003

Information on nutrient requirements and forage intake of fallow weaner deer is required for the development of feeding strategies during the year. An experiment was conducted in which 60 fallow weaner deer (grazing on medic and ryegrass based pastures) were supplemented with a concentrated diet at three levels. The diet contained 2% minerals, 30% lupin and 68% barley grain. Twelve deer from each treatment were dosed with commercial alkane capsules in May, June, July, September and October to predict nutrient intake. The relationships between body weight gain and intake of metabolisable energy and crude protein were established using a general linear models analysis. Dry matter intake from pastures ranged from 0.137 kg to 0.304 kg in May and June and increased to 1.2 kg in October. Nutrient intake from pastures was strongly influenced by amount of supplementary feed and gender. Digestible energy intake from pastures was 1.3, 3.8 and 6.1 MJ/day higher for males than females in July, August and October, respectively. The protein and energy intake was strongly correlated with body weight gain. The energy requirement for maintenance were 7.3, 8.2, 10.2, 10.2 and 10.7 MJ DE/day and the DE required for each kg body weight gain were 19, 18, 29, 34 and 49 MJ in May, June, August and October, respectively. The protein requirement for maintenance was 12.2, 12.6, 15.0, 11.4 and $8.5g/W^{0.75}$ in May, June, July, August and October, respectively. The nutrient requirement defined from this study can be used to assist farmers to explore the possible pasture and stock management practices under southern Australian conditions. However, further research is required to develop rapid and cheap methods for estimating dry matter intake, nutritive value of pastures and to quantify the potential growth rate of fallow deer in southern Australia.

• Asian-Australasian Journal of Animal Sciences
• /
• v.23 no.8
• /
• pp.1089-1095
• /
• 2010

Four biological candidate genes, natural resistance associated macrophage protein 1 (SLC11A1 or NRAMP), prosaposin (PSAP), interferon Gamma (IFNG), and toll-like receptor 4 (TLR4), were examined to identify single nucleotide polymorphisms (SNP) and associations of the SNP with antibody response kinetics in hens. An $F_2$ population was produced by mating $G_0$ highly inbred (<99%) males of two MHC-congenic Fayoumi lines with highly inbred Leghorn hens. The $F_2$ hens (n = 158) were injected twice with SRBC and whole, fixed Brucella abortus (BA). Blood samples were obtained before each immunization, at 7 d after primary immunization, and at several time points after secondary immunization. Minimum titers (Ymin) and the time needed to reach them (Tmin), and maximum (Ymax) titers and the time needed to reach them (Tmax), were estimated from the seven post-secondary immunization titers using a nonlinear regression model. The $F_2$ hens were genotyped for the four candidate genes by using PCR-RFLP for one SNP per gene, which identified the parental allele. General linear models were used to test associations of SNP genotypes with antibody response parameters and BW measured at 4 ages. The IFNG SNP was highly significantly (p<0.0125) associated with primary response to SRBC, Tmin to BA, Ymin to BA, and 12-week BW. The current study demonstrated that the novel IFNG promoter SNP was associated with antibody kinetics for BA and SRBC in laying hens, and also with BW, suggesting that this cytokine may play a pivotal role in the relationship between immune function and growth.