• Journal of Family Resource Management and Policy Review
• /
• v.22 no.1
• /
• pp.63-78
• /
• 2018

This study examined the insolvency of debtors using multiple-indicator approaches and compared the outcomes across income levels with the 2016 'Household Financial and Welfare Survey'. This study used (1) the total debt to total assets ratio (DTA), (2) the total debt service ratio (DSR), and (3) the Household Default Risk Index (HDRI) recently developed by the Bank of Korea. Households in the lowest income quintile were more likely to be insolvent than any other income group. Demographics, such as age and gender of the household head, and most of the financial variables significantly increased the likelihood of insolvency based on the DTA. The number of household members and job status increased the likelihood of insolvency based on the DSR. Also, age, gender of the household head, and most of the financial variables increased the likelihood of household insolvency based on the HDRI after controlling for other demographics and financial variables.

• Journal of Family Resource Management and Policy Review
• /
• v.23 no.2
• /
• pp.99-115
• /
• 2019

This study examined the insolvency likelihood of young debtors from the 2018 Household Financial and Welfare Survey. This study used the Household Default Risk Index (HDRI), which considers the ratio of total debt to total assets (DTA), and a total debt service ratio (DSR) to examine the insolvency level of debtors. The descriptive analyses showed no difference in frequency of households with a high probability of insolvency between those less than 35 years of age and those over 35 years of age. However, the median HDRI value for those less than 35 years of age was higher than those over 35 years of age. The multivariate analyses indicated that educational expenses for young Korean debtors was a factor that increased their probability of insolvency, while income was the only variable that decreased their insolvency likelihood.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.5
• /
• pp.94-102
• /
• 1998

In this paper, we proposed a dynamic selection scheme of advnaced prediction mode(DAPM), which reduces computational cost and improves coding efficiency. We can select the mode between default prediction mode (DPM) and advanced prediction mode (APM) according to motion componenets in a frame dynamically. For this purpose, we defined error distribution of motion estimation (EDME) as sum of absolute difference(SAD) for each searching points. This distribution region is divided to four subregions. We calculate minimum values in each subregions and then, we determine whether block motion estimation is performed or not depending on the results. As a result, we reduced computational complexity to 30% without degradation of image quality compared to fixed APM(FAPM) by selecting DPM for linear movement.

• Journal of information and communication convergence engineering
• /
• v.19 no.4
• /
• pp.228-233
• /
• 2021

In this study, we analyze the credit information (loan, delinquency information, etc.) of individual business owners to generate voluminous training data to establish a bankruptcy prediction model through a partial synthetic training technique. Furthermore, we evaluate the prediction performance of the newly generated data compared to the actual data. When using conditional tabular generative adversarial networks (CTGAN)-based training data generated by the experimental results (a logistic regression task), the recall is improved by 1.75 times compared to that obtained using the actual data. The probability that both the actual and generated data are sampled over an identical distribution is verified to be much higher than 80%. Providing artificial intelligence training data through data synthesis in the fields of credit rating and default risk prediction of individual businesses, which have not been relatively active in research, promotes further in-depth research efforts focused on utilizing such methods.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.42 no.6
• /
• pp.63-74
• /
• 2005

In Electronic Commerce, the latest most of the personalized recommender systems have applied to the collaborative filtering technique. This method calculates the weight of similarity among users who have a similar preference degree in order to predict and recommend the item which hits to propensity of users. In this case, we commonly use Pearson Correlation Coefficient. However, this method is feasible to calculate a correlation if only there are the items that two users evaluated a preference degree in common. Accordingly, the accuracy of prediction falls. The weight of similarity can affect not only the case which predicts the item which hits to propensity of users, but also the performance of the personalized recommender system. In this study, we verify the improvement of the prediction accuracy through an experiment after observing the rule of the weight of similarity applying Vector similarity, Entropy, Inverse user frequency, and Default voting of Information Retrieval field. The result shows that the method combining the weight of similarity using the Entropy with Default voting got the most efficient performance.

• Journal of the Korean Society of Urban Environment
• /
• v.18 no.4
• /
• pp.401-408
• /
• 2018

EQPS (Effluent Quality Prediction System, Dynamita, France) was applied to analyze the appropriateness of the design of a bioreactor in A sewage treatment plant. A sewage treatment plant was designed by setting the design concentration of the secondary clarifier effluent to total nitrogen and total phosphorus, 10 mg/L and 1.8 mg/L, respectively, in order to comply with the target water quality at the level of the hydrophilic water. The retention time of the 4-stage BNR reactor was 9.6 hours, which was 0.5 for the pre-anoxic tank, 1.0 for the anaerobic tank, 2.9 for the anoxic tank, and 5.2 hours for the aerobic tank. As a result of the modeling of the winter season, the retention time of the anaerobic tank was increased by 0.2 hours in order to satisfy the target water quality of the hydrophilic water level. The default coefficients of the one step nitrification denitrification model proposed by the software manufacturer were used to exclude distortion of the modeling results. Since the process modeling generally presents optimal conditions, the retention time of the 4-stage BNR should be increased to 9.8 hours considering the bioreactor margin. The accurate use of process modeling in the design stage of the sewage treatment plant is a way to ensure the stability of the treatment performance and efficiency after construction of the sewage treatment plant.

• Smart Media Journal
• /
• v.13 no.4
• /
• pp.9-15
• /
• 2024

Tree-based algorithms have been the dominant methods used build a prediction model for tabular data. This also includes personal credit data. However, they are limited to compatibility with categorical and numerical data only, and also do not capture information of the relationship between other features. In this work, we proposed an ensemble model using the Transformer architecture that includes text features and harness the self-attention mechanism to tackle the feature relationships limitation. We describe a text formatter module, that converts the original tabular data into sentence data that is fed into FinBERT along with other text features. Furthermore, we employed FT-Transformer that train with the original tabular data. We evaluate this multi-modal approach with two popular tree-based algorithms known as, Random Forest and Extreme Gradient Boosting, XGBoost and TabTransformer. Our proposed method shows superior Default Recall, F1 score and AUC results across two public data sets. Our results are significant for financial institutions to reduce the risk of financial loss regarding defaulters.

• Journal of Multimedia Information System
• /
• v.7 no.2
• /
• pp.141-146
• /
• 2020

Cine-Hangeul is a program that can predict the running time of a movie based on the screenplay before production. This paper seeks to verify the prediction reporting function of Cine-Hangeul, which is the standard Korean screenplay format. Moreover, this paper presents a method to increase the accuracy of the Cine-Hangeul reporting function. The objective of this paper is to offer a correction method based on scientific evidence because the current Cine-Hangeul reporting function has many errors. The verification process for five scenarios and movies confirmed that the default setting value of Cine- Hangeul's screening time prediction reporting was many errors. Cine-Hangeul analyzes the amount of textual information to predict the time of the scene and the time of the dialogue and helps predict the total time of the movie. Therefore, if a certain amount of text information is not available, the accuracy is unreliable. The current Cine-Hangeul prediction report confirms that the efficiency is high when the scenario volume is about 90 to 100 pages. As a result, prediction of screening time by Cine-Hangeul, a Korean scenario standard format program, confirmed the verification that it could secure the same level of reliability as the actual screening time by correcting the reporting settings. This verification also affirms that when applying about 50 percent of the basic set of screening time reporting, it is almost identical to the screening time.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2002.05a
• /
• pp.1059-1065
• /
• 2002

주가자료를 활용한 부도예측모형인 KMV EDF 모형을 기반으로 일별주가자료와 기업재무자료를 이용하여, 모형에 필요한 적절한 모수를 찾고 모델링을 하였으며, 적절성을 검증했다. 그리고, 기존의 연구에 따라 월평균주가자료를 이용한 경우, 모형에 왜곡이 가해질 수 있다는 점을 지적했다. 또한, 민감도 분석을 통하여 본 모형의 부도예측값에 미치는 주요한 검증하고, 실용적으로 사용할 수 있는 간단한 민감도분석 Tool을 설계하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.5B
• /
• pp.369-380
• /
• 2003

In the CDMA wireless link, frame errors are correlated and burst because of fading. The implementation ability of RLP error recovery is dependent on the correlated frame errors. The (1,2,3) retransmission scheme, which is recommended as default in IS-707, is not adapted in high frame loss regime with strong correlations. By using the modified error recovery method, where the total number of retransmission attempts is the same and the retransmission is increased, the proposed retransmission scheme can efficiently recover frame errors than the (1,2,3) retransmission scheme. Since the modified scheme has longer transmission delay due to the increase of retransmission round, we propose the algorithm of retransmission failure prediction to improve the modified error recovery scheme. We simulate the modified error recovery scheme applying our algorithm and compare two schemes. (i.e. default scheme by IS-707 and modified scheme) not applying our algorithm. In the result, we show TCP performance improvement is better than default scheme by IS-707.