• Journal of Intelligence and Information Systems
• /
• v.26 no.2
• /
• 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.

• The Korean Journal of Nuclear Medicine Technology
• /
• v.20 no.2
• /
• pp.14-20
• /
• 2016

Purpose Tongue cancer is 1.8% of all cancer tumors occur in the tongue, it is known that the high incidence enough to account for 75% of oral cancer conducted a PET / CT examination for early diagnosis, metastasis, staging, etc. and. Tongue when PET / CT scan of a cancer patient and a Torso taken to close mouth lesions if the condition was caused due to the overlapping or corresponding artifacts are not clearly observed. The purpose of this study is to evaluate the changes that occur during PET / CT scan with open mouth and its usefulness under. Materials and Methods From June 2015 to March 2016 complained of herein by May 21 had received a diagnosis of tongue cancer underwent PET / CT scan patients were treated with a target (16 males, 5 female). The first was taken to close mouth Torso state, it was taken to add 1 bed open mouth condition. Tumor (T), measuring the Normal Tongue (NT), Lymph Node (LN) standard intake coefficient by setting a region of interest in the (standardized uptake value, SUV) SUVmean, the average value was measured SUVmax, drawn to each region of interest 3 times and Background (Carotid artery) was out of the SUV. In Chapter 3 of the slice to the tumor clearly visible by setting the region of interest to measure the change Tumor size was calculated average value. Gross Image resolution assessment were analyzed statistically through were divided into 1-5 points by the Radiation 7 people in 2, more than five years worked in specialized nuclear medicine compare to proceed with the blind test nonparametric test (wilcoxon signed rank test). (SPSS ver.18) Results $SUV_{mean}$ T's were in close mouth $5.01{\pm}2.70$ with open mouth $5.48{\pm}2.88$ (P<0.05), $SUV_{max}$ were respectively $8.78{\pm}5.55$ and $9.70{\pm}5.99$ (P<0.05). $SUV_{mean}$ in the NT were respectively $0.43{\pm}0.30$ and $0.34{\pm}0.24$ (P=0.20), $SUV_{max}$ was $0.56{\pm}0.34$ and $0.45{\pm}0.25$ (P=0.204). LN $SUV_{mean}$ were respectively $1.62{\pm}1.43$ and $1.69{\pm}1.49$ (P=0.161), $SUV_{mean}$ was $2.09{\pm}1.88$ and $1.99{\pm}1.74$ (P=0.131). Tumor size change is close mouth $4.96{\pm}4.66cm^2$ $5.33{\pm}4.64cm^2$ with 7.45% increase was (P<0.05), gross image resolution evaluation is $2.87{\pm}0.73$, $3.77{\pm}0.68$ with open mouth examinations 30.5% increase was (P<0.05). Conclusion Tumor SUV on the changes that had an increase in open mouth during inspection, the normal tongue and lymph node, but there was no significant difference in the change slightly. It is also one open mouth PET / CT scan will provide improved image to all patients with tongue cancer, but it could be confirmed that similar overall through the blind test, or tumor size changes and showing a high resolution image. It can be the perfect alternative method for problems that occur when the close mouth Open mouth PET / CT scan, but is believed to be through the open mouth to observe the boundary of overlapping or tumor of the oral cavity other structures a little more clearly. Tongue cancer patients how to recommend that the shooting further open mouth PET / CT.