• 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.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.175-197
• /
• 2021

Recently, with the development of deep learning technology, research on unstructured data analysis is being actively conducted, and it is showing remarkable results in various fields such as classification, summary, and generation. Among various text analysis fields, text classification is the most widely used technology in academia and industry. Text classification includes binary class classification with one label among two classes, multi-class classification with one label among several classes, and multi-label classification with multiple labels among several classes. In particular, multi-label classification requires a different training method from binary class classification and multi-class classification because of the characteristic of having multiple labels. In addition, since the number of labels to be predicted increases as the number of labels and classes increases, there is a limitation in that performance improvement is difficult due to an increase in prediction difficulty. To overcome these limitations, (i) compressing the initially given high-dimensional label space into a low-dimensional latent label space, (ii) after performing training to predict the compressed label, (iii) restoring the predicted label to the high-dimensional original label space, research on label embedding is being actively conducted. Typical label embedding techniques include Principal Label Space Transformation (PLST), Multi-Label Classification via Boolean Matrix Decomposition (MLC-BMaD), and Bayesian Multi-Label Compressed Sensing (BML-CS). However, since these techniques consider only the linear relationship between labels or compress the labels by random transformation, it is difficult to understand the non-linear relationship between labels, so there is a limitation in that it is not possible to create a latent label space sufficiently containing the information of the original label. Recently, there have been increasing attempts to improve performance by applying deep learning technology to label embedding. Label embedding using an autoencoder, a deep learning model that is effective for data compression and restoration, is representative. However, the traditional autoencoder-based label embedding has a limitation in that a large amount of information loss occurs when compressing a high-dimensional label space having a myriad of classes into a low-dimensional latent label space. This can be found in the gradient loss problem that occurs in the backpropagation process of learning. To solve this problem, skip connection was devised, and by adding the input of the layer to the output to prevent gradient loss during backpropagation, efficient learning is possible even when the layer is deep. Skip connection is mainly used for image feature extraction in convolutional neural networks, but studies using skip connection in autoencoder or label embedding process are still lacking. Therefore, in this study, we propose an autoencoder-based label embedding methodology in which skip connections are added to each of the encoder and decoder to form a low-dimensional latent label space that reflects the information of the high-dimensional label space well. In addition, the proposed methodology was applied to actual paper keywords to derive the high-dimensional keyword label space and the low-dimensional latent label space. Using this, we conducted an experiment to predict the compressed keyword vector existing in the latent label space from the paper abstract and to evaluate the multi-label classification by restoring the predicted keyword vector back to the original label space. As a result, the accuracy, precision, recall, and F1 score used as performance indicators showed far superior performance in multi-label classification based on the proposed methodology compared to traditional multi-label classification methods. This can be seen that the low-dimensional latent label space derived through the proposed methodology well reflected the information of the high-dimensional label space, which ultimately led to the improvement of the performance of the multi-label classification itself. In addition, the utility of the proposed methodology was identified by comparing the performance of the proposed methodology according to the domain characteristics and the number of dimensions of the latent label space.

• Radiation Oncology Journal
• /
• v.16 no.4
• /
• pp.409-423
• /
• 1998

Purpose : This retrospective study was tried to evaluate the clinical characteristics of patients, patterns of failure, survival rates, prognostic factors affecting survival, and treatment related toxicities when non-small cell lung cancer patients was treated by definitive radiotherapy alone or combined with chemotherapy. Materials and Methods : We evaluated the treatment results of 70 patients who were treated by definitive radiation therapy for non-small cell lung cancer at the Department of Radiation Oncology, Ewha Womans University Hospital, between March 1982 and April 1996. The number of patients of each stage was 2 in stage I, 6 in stage II, 30 in stage III-A, 29 in stage III-B, 3 in stage IV. Radiation therapy was administered by 6 MV linear accelerator and daily dose was 1.8-2.0 Gy and total radiation dose was ranged from 50.4 Gy to 72.0 Gy with median dose 59.4 Gy. Thirty four patients was treated with combined therapy with neoadjuvant or concurrent chemotherapy and radiotherapy, and most of them were administered with the multi-drug combined chemotherapy including etoposide and cisplatin. The survival rate was calculated with the Kaplan-Meier methods. Results : The overall 1-year, 2-year, and 3-year survival rates were 63$\%$, 29$\%$, and 26$\%$, respectively. The median survival time of all patients was 17 months. The disease-free survival rate for 1-year and 2-year were 23$\%$ and 16$\%$, respectively. The overall 1-year survival rates according to the stage was 100$\%$ for stage I, 80$\%$ for stage II, 61$\%$ for stage III, and 50$\%$ for stage IV. The overall 1-year 2-year, and 3-year survival rates for stage III patients only were 61$\%$, 23$\%$, and 20$\%$, respectively. The median survival time of stage III patients only was 15 months. The complete response rates by radiation therapy was 10$\%$ and partial response rate was 50$\%$. Thirty patients (43$\%$) among 70 patients assessed local control at initial 3 months follow-up duration. Twenty four (80$\%$) of these 30 Patients was possible to evaluate the pattern of failure after achievement of local control. And then, treatment failure occured in 14 patients (58$\%$): local relapse in 6 patients (43$\%$), distant metastasis in 6 patients (43$\%$) and local relapse with distant metastasis in 2 patients (14$\%$). Therefore, 10 patients (23$\%$) were controlled of disease of primary site with or without distant metastases. Twenty three patients (46$\%$) among 50 patients who were possible to follow-up had distant metastasis. The overall 1-year survival rate according to the treatment modalities was 59$\%$ in radiotherapy alone and 66$\%$ in chemoirradiation group. The overall 1-year survival rates for stage III patients only was 51$\%$ in radiotherapy alone and 68$\%$ in chemoirradiation group which was significant different. The significant prognostic factors affecting survival rate were the stage and the achievement of local control for all patients at univariate- analysis. Use of neoadjuvant or concurrent chemotherapy, use of chemotherapy and the achievement of local control for stage III patients only were also prognostic factors. The stage, pretreatment performance status, use of neoadjuvant or concurrent chemotherapy, total radiation dose and the achievement of local control were significant at multivariate analysis. The treatment-related toxicities were esophagitis, radiation pneunonitis, hematologic toxicity and dermatitis, which were spontaneously improved, but 2 patients were died with radiation pneumonitis. Conclusion : The conventional radiation therapy was not sufficient therapy for achievement of long-term survival in locally advanced non-small cell lung cancer. Therefore, aggressive treatment including the addition of appropriate chemotherapeutic drug to decrease distant metastasis and preoperative radiotherapy combined with surgery, hyperfractionation radiotherapy or 3-D conformal radiation therapy for increase local control are needed.

• Asia pacific journal of information systems
• /
• v.19 no.2
• /
• pp.139-155
• /
• 2009

For the last few decades, many studies have tried to explore and unveil venture companies' success factors and unique features in order to identify the sources of such companies' competitive advantages over their rivals. Such venture companies have shown tendency to give high returns for investors generally making the best use of information technology. For this reason, many venture companies are keen on attracting avid investors' attention. Investors generally make their investment decisions by carefully examining the evaluation criteria of the alternatives. To them, credit rating information provided by international rating agencies, such as Standard and Poor's, Moody's and Fitch is crucial source as to such pivotal concerns as companies stability, growth, and risk status. But these types of information are generated only for the companies issuing corporate bonds, not venture companies. Therefore, this study proposes a method for evaluating venture businesses by presenting our recent empirical results using financial data of Korean venture companies listed on KOSDAQ in Korea exchange. In addition, this paper used multi-class SVM for the prediction of DEA-based efficiency rating for venture businesses, which was derived from our proposed method. Our approach sheds light on ways to locate efficient companies generating high level of profits. Above all, in determining effective ways to evaluate a venture firm's efficiency, it is important to understand the major contributing factors of such efficiency. Therefore, this paper is constructed on the basis of following two ideas to classify which companies are more efficient venture companies: i) making DEA based multi-class rating for sample companies and ii) developing multi-class SVM-based efficiency prediction model for classifying all companies. First, the Data Envelopment Analysis(DEA) is a non-parametric multiple input-output efficiency technique that measures the relative efficiency of decision making units(DMUs) using a linear programming based model. It is non-parametric because it requires no assumption on the shape or parameters of the underlying production function. DEA has been already widely applied for evaluating the relative efficiency of DMUs. Recently, a number of DEA based studies have evaluated the efficiency of various types of companies, such as internet companies and venture companies. It has been also applied to corporate credit ratings. In this study we utilized DEA for sorting venture companies by efficiency based ratings. The Support Vector Machine(SVM), on the other hand, is a popular technique for solving data classification problems. In this paper, we employed SVM to classify the efficiency ratings in IT venture companies according to the results of DEA. The SVM method was first developed by Vapnik (1995). As one of many machine learning techniques, SVM is based on a statistical theory. Thus far, the method has shown good performances especially in generalizing capacity in classification tasks, resulting in numerous applications in many areas of business, SVM is basically the algorithm that finds the maximum margin hyperplane, which is the maximum separation between classes. According to this method, support vectors are the closest to the maximum margin hyperplane. If it is impossible to classify, we can use the kernel function. In the case of nonlinear class boundaries, we can transform the inputs into a high-dimensional feature space, This is the original input space and is mapped into a high-dimensional dot-product space. Many studies applied SVM to the prediction of bankruptcy, the forecast a financial time series, and the problem of estimating credit rating, In this study we employed SVM for developing data mining-based efficiency prediction model. We used the Gaussian radial function as a kernel function of SVM. In multi-class SVM, we adopted one-against-one approach between binary classification method and two all-together methods, proposed by Weston and Watkins(1999) and Crammer and Singer(2000), respectively. In this research, we used corporate information of 154 companies listed on KOSDAQ market in Korea exchange. We obtained companies' financial information of 2005 from the KIS(Korea Information Service, Inc.). Using this data, we made multi-class rating with DEA efficiency and built multi-class prediction model based data mining. Among three manners of multi-classification, the hit ratio of the Weston and Watkins method is the best in the test data set. In multi classification problems as efficiency ratings of venture business, it is very useful for investors to know the class with errors, one class difference, when it is difficult to find out the accurate class in the actual market. So we presented accuracy results within 1-class errors, and the Weston and Watkins method showed 85.7% accuracy in our test samples. We conclude that the DEA based multi-class approach in venture business generates more information than the binary classification problem, notwithstanding its efficiency level. We believe this model can help investors in decision making as it provides a reliably tool to evaluate venture companies in the financial domain. For the future research, we perceive the need to enhance such areas as the variable selection process, the parameter selection of kernel function, the generalization, and the sample size of multi-class.

• Radiation Oncology Journal
• /
• v.18 no.4
• /
• pp.244-250
• /
• 2000

Purpose : To see the relationship between the response to chemotherapy and the final outcome of neoadiuvant chemotherapy and radiotherapy in patients with mocanry advanced hypopharyngeal cancer. Methods and Materials :A retrospective analysis was done for thirty-two patients with locally advanced hypopharyngeal cancer treated in the Seoul National University Hospital with neoadiuvant chemotherapy and radiotherapy from August 1979 to July 1997. The patients were treated with Co-60 teletherapy unit or 4MV or 6MV photon beam produced by linear accelerator. Daily fractionation was 1.75 to 2 Gy, delivered five times a week. Total dose ranged from 60.8 Gy to 73.8 Gy. Twenty-nine patients received continuous infusion of cisplatin and 5-FU. Other patients were treated with cisplatin combined with bleomycin or vinblastin. Twenty-four (75$\%$) patients received all three prescribed cycles of chemotherapy delivered three weeks apart. Six patients received two cycles, and two patients received only one cycle. Results :The overall 2-year and 5-year survival rates are 65.6$\%$ and 43.0$\%$, respectively. 5-year local control rate is 34$\%$. Organ preservation for more than five years is achieved in 12 patients (38$\%$). After neoadjuvant chemotherapy, 24 patients achieved more than partial remission (PR): the response rate was 75$\%$ (24/32). Five patients had complete remission (CR), 19 patients PR, and 8 patients no response (NR). Among the 19 patients who had PR to chemotherapy, 8 patients achieved CR after radiotherapy. Among the 8 non-responders to chemotherapy, 2 patients achieved CR, and 6 patients achieved PR after radiotherapy. There was no non-responder after radiotherapy. The overall survival rates were 60$\%$ for CR to chemotherapy group, 35.1$\%$ for PR to chemotherapy group, and 50$\%$ for NR to chemotherapy group, respectively (p=0.93). There were significant difference in five-year overall survival rates between the patients with CR and PR after neoadjuvant chemotherapy and radiotherapy (73.3$\%$ vs. 14.7$\%$, p<0.01). The prognostic factor affecting overall survival was the response to overall treatment (CR vs. PR, p<0.01). Conclusion :In this study, there were only five patients who achieved CR after neoadiuvant chemotherapy. Therefore the difference of overall survival rates between CR and PR to chemotherapy group was not statistically significant. Only the response to chemo-radiotherapy was the most important prognostic factor. There needs to be more effort to improve CR rate of neoadjuvant chemotherapy and consideration for future use of concurrent chemoradiotherapy.

• Radiation Oncology Journal
• /
• v.16 no.1
• /
• pp.71-79
• /
• 1998

Purpose : The objective of this study is to introduce our installation of a non-commercial 3D Planning system, Plunc and confirm it's clinical applicability in various treatment situations. Materials and Methods : We obtained source codes of Plunc, offered by University of North Carolina and installed them on a Pentium Pro 200MHz (128MB RAM, Millenium VGA) with Linux operating system. To examine accuracy of dose distributions calculated by Plunc, we input beam data of 6MV Photon of our linear accelerator(Siemens MXE 6740) including tissue-maximum ratio, scatter-maximum ratio, attenuation coefficients and shapes of wedge filters. After then, we compared values of dose distributions(Percent depth dose; PDD, dose profiles with and without wedge filters, oblique incident beam, and dose distributions under air-gap) calculated by Plunc with measured values. Results : Plunc operated in almost real time except spending about 10 seconds in full volume dose distribution and dose-volume histogram(DVH) on the PC described above. As compared with measurements for irradiations of 90-cm 550 and 10-cm depth isocenter, the PDD curves calculated by Plunc did not exceed $1\%$ of inaccuracies except buildup region. For dose profiles with and without wedge filter, the calculated ones are accurate within $2\%$ except low-dose region outside irradiations where Plunc showed $5\%$ of dose reduction. For the oblique incident beam, it showed a good agreement except low dose region below $30\%$ of isocenter dose. In the case of dose distribution under air-gap, there was $5\%$ errors of the central-axis dose. Conclusion : By comparing photon dose calculations using the Plunc with measurements, we confirmed that Plunc showed acceptable accuracies about $2-5\%$ in typical treatment situations which was comparable to commercial planning systems using correction-based a1gorithms. Plunc does not have a function for electron beam planning up to the present. However, it is possible to implement electron dose calculation modules or more accurate photon dose calculation into the Plunc system. Plunc is shown to be useful to clear many limitations of 2D planning systems in clinics where a commercial 3D planning system is not available.

• Clinical and Experimental Pediatrics
• /
• v.50 no.6
• /
• pp.549-555
• /
• 2007

Purpose : Pulse wave velocity (PWV) and ankle brachial index (ABI) are simple, non-invasive methods to assess arterial stiffness. These parameters are also known to be closely related to cardiovascular risk factors and diseases. The purposes of this study were to measure blood pressure, PWV, ABI in healthy Korean adolescents, set up their normal values and assess their correlations. Methods : Three hundred ninety two healthy adolescents (213 boys and 179 girls) underwent measurement of brachial ankle pulse wave velocity (baPWV), ABI, body mass index(BMI) and blood pressure from four extremities. Linear regression analysis was performed to reveal the correlations between PWV, ABI and independent variables. Results : Blood pressure and PWV were significantly higher in all extremities in males compared to females. Blood pressure of both brachial and ankle showed positive correlation with body weight, height, and BMI, whereas ABI showed no correlation with any of these indices. Conclusion : Blood pressure increases as body weight, height and BMI increases. PWV shows positive correlation with blood pressure. It will be helpful to predict the risks of cardiovascular diseases in adolescents.

• Radiation Oncology Journal
• /
• v.28 no.1
• /
• pp.32-38
• /
• 2010

Purpose: Although radiation-induced fibrosis is one of the common sequelae occurring after irradiation of skin and soft tissues, the treatment methods are not well standardized. This study aimed to establish the skin fibrosis mouse model by fractionated radiation for the further mechanism studies or testing the efficacy of therapeutic candidates. Materials and Methods: The right hind limbs of BALB/c mice received two fractions of 20 Gy using a therapeutic linear accelerator. Early skin damages were scored and tissue fibrosis was assessed by the measurement of a leg extension. Morphological changes were assessed by H&E staining and by Masson's Trichrome staining. TGF-${\beta}1$ expression from soft tissues was also detected by immunohistochemistry and PCR. Results: Two fractions of 20 Gy irradiation were demonstrated as being enough to induce early skin damage effects such as erythema, mild skin dryness, dry and wet desquamation within several weeks of radiation. After 13 weeks of irradiation, the average radiation-induced leg contraction was $11.1{\pm}6.2mm$. Morphologic changes in irradiated skin biopsies exhibited disorganized collagen and extracellular matrix fibers, as well as the accumulation of myofibroblasts compared to the non-irradiated skin. Moreover, TGF-${\beta}1$ expression in tissue was increased by radiation. Conclusion: These results show that two fractions of 20 Gy irradiation can induce skin fibrosis in BALB/c mice accompanied by other common characteristics of skin damages. This animal model can be a useful tool for studying skin fibrosis induced by radiation.

• Korean Journal of Psychosomatic Medicine
• /
• v.31 no.2
• /
• pp.108-117
• /
• 2023

Objectives : Somatic symptom disorder (SSD) is characterized by the manifestation of a variety of physical symptoms, but little is known about differences in autonomic nervous system activity according to symptom severity, especially within patient groups. In this study, we examined differences in heart rate variability (HRV) across symptom severity in a group of SSD patients to analyze a representative marker of autonomic nervous system changes by symptoms severity. Methods : Medical records were retrospectively reviewed for patients who were diagnosed with SSD based on DSM-5 from September 18, 2020 to October 29, 2021. We applied inverse probability of treatment weighting (IPTW) methods to generate more homogeneous comparisons in HRV parameters by correcting for selection biases due to sociodemographic and clinical characteristic differences between groups. Results : There were statistically significant correlations between the somatic symptom severity and LF (nu), HF (nu), LF/HF, as well as SD1/SD2 and Alpha1/Alpha2. After IPTW estimation, the mild to moderate group was corrected to 27 (53.0%) and the severe group to 24 (47.0%), and homogeneity was achieved as the differences in demographic and clinical characteristics were not significant. The analysis of inverse probability weighted regression adjustment model showed that the severe group was associated with significantly lower RMSSD (β=-0.70, p=0.003) and pNN20 (β=-1.04, p=0.019) in the time domain and higher LF (nu) (β=0.29, p<0.001), lower HF (nu) (β=-0.29, p<0.001), higher LF/HF (β=1.41, p=0.001), and in the nonlinear domain, significant differences were tested for SampEn15 (β=-0.35, p=0.014), SD1/SD2 (β=-0.68, p<0.001), and Alpha1/Alpha2 (ß=0.43, p=0.001). Conclusions : These results suggest that differences in HRV parameters by SSD severity were showed in the time, frequency and nonlinear domains, specific parameters demonstrating significantly higher sympathetic nerve activity and reduced ability of the parasympathetic nervous system in SSD patients with severe symptoms.

• Radiation Oncology Journal
• /
• v.21 no.4
• /
• pp.253-260
• /
• 2003

Purpose: No general consensus has been reached regarding the necessity of postoperative radiation therapy (PORT) and the optimal techniques of its application for patients with chest wall invasion (pT3cw) and node negative (NO) non-small cell lung cancer (NSCLC). We retrospectively analyzed the PT3cwN0 NSCLC patients who received PORT because of presumed inadequate resection margin on surgical findings. Materials and Methods: From Aug. 1994 till June 2000, 21 pT3cwN0 NSCLC patients received PORT at Samsung Medical Center; all of whom underwent curative on-bloc resection of the primary tumor plus the chest wall and regional lymph node dissection. PORT was typically stalled 3 to 4 weeks after operation using 6 or 10 MV X-rays from a linear accelerator. The radiation target volume was confined to the tumor bed plus the immediate adjacent tissue, and no regional lymphatics were included. The planned radiation dose was 54 Gy by conventional fractionation schedule. The survival rates were calculated and the failure patterns analyzed. Results: Overall survival, disease-free survival, loco-regional recurrence-free survival, and distant metastases-free survival rates at 5 years were 38.8$\%$, 45.5$\%$, 90.2$\%$, and 48.1$\%$, respectively. Eleven patients experienced treatment failure: six with distant metastases, three with intra-thoracic failures, and two with combined distant and intra-thoracic failures. Among the five patients with intra-thoracic failures, two had pleural seeding, two had in-field local failures, and only one had regional lymphatic failure in the mediastinum. No patients suffered from acute and late radiation side effects of RTOG grade 3 or higher. Conclusion: The strategy of adding PORT to surgery to improve the probability, not only of local control but also of survival, was justified, considering that local control was the most important component in the successful treatment of pT3cw NSCLC patients, especially when the resection margin was not adequate. The incidence and the severity of the acute and late side effects of PORT were markedly reduced, which contributed to improving the patients' qualify of life both during and after PORT, without increasing the risk of regional failures by eliminating the regional lymphatics from the radiation target volume.