• Proceedings of the Korean Institute of Building Construction Conference
• /
• 2022.11a
• /
• pp.99-100
• /
• 2022

As the danger of exposure to the asbestos has been revealed, the importance of demolition asbestos in existing buildings has been raised. Extensive body of study has been conducted to evaluate the risk of demolition asbestos, but there were confined types of variables caused by not reflecting categorical information and limitations in collecting quantitative information. Thus, this study aims to derive a model that predicts the risk in workplace of demolition asbestos by collecting categorical and continuous variables. For this purpose, categorical and continuous variables were collected from asbestos demolition reports, and the risk assessment score was set as the dependent variable. In this study, the influence of each variable was identified using logistic regression, and the risk prediction model methodologies were compared through decision tree regression and artificial neural network. As a result, a conditional risk prediction model was derived to evaluate the risk of demolition asbestos, and this model is expected to be used to ensure the safety of asbestos demolition workers.

• Proceedings of the Korean Society of Disaster Information Conference
• /
• 2017.11a
• /
• pp.339-340
• /
• 2017

화학공장의 누출사고는 초기에 적절히 대응하지 못할 경우 화재 폭발과 같은 2차 3차의 복합재난사고로 확산될 위험성이 매우 높다. 이러한 이유로 누출사고 발생 초기에 누출이 발생한 지점을 신속히 파악하여 현장안전요원에게 알림으로써, 보다 체계적이고 효율적인 초기대응을 가능하게 하여, 사고피해를 완화시킬 수 있는 통합적인 누출사고 대응시스템 구축은 매우 중요하다고 할 수 있다. 본 연구에서는, 통합적인 누출사고 대응시스템 구축을 위한 선행연구로, 딥러닝 기반의 누출원추적 모델 개발을 제안한다. 여수에 위치한 실제 화학공장을 대상으로 누출사고 시나리오에 대한 Computational Fluid Dynamics (CFD) 시뮬레이션을 진행한 뒤, 화학공장 경계면에 배치된 각 센서별 위치에서의 농도, 풍향 그리고 풍속데이터를 추출하고, 센서 좌표를 추가하여 인공신경망을 학습시켰다. 학습된 모델은 40개의 누출후보군에 대해 학습에 사용되지 않은 상황들에서도 75.43%의 정확도로 누출이 일어난 지점을 실시간 예측해냄을 확인하였다. 또한 누출지점 예측이 일치하지 않은 경우도, 예측된 지점이 실제 누출이 일어난 지점과 물리적으로 매우 인접함을 확인함으로써 제안된 모델을 실제 현장에 적용할시 기대되는 효과는 더 클 것으로 판단하였다.

• Journal of Intelligence and Information Systems
• /
• v.29 no.2
• /
• pp.241-265
• /
• 2023

A corporate insolvency prediction model serves as a vital tool for objectively monitoring the financial condition of companies. It enables timely warnings, facilitates responsive actions, and supports the formulation of effective management strategies to mitigate bankruptcy risks and enhance performance. Investors and financial institutions utilize default prediction models to minimize financial losses. As the interest in utilizing artificial intelligence (AI) technology for corporate insolvency prediction grows, extensive research has been conducted in this domain. However, there is an increasing demand for explainable AI models in corporate insolvency prediction, emphasizing interpretability and reliability. The SHAP (SHapley Additive exPlanations) technique has gained significant popularity and has demonstrated strong performance in various applications. Nonetheless, it has limitations such as computational cost, processing time, and scalability concerns based on the number of variables. This study introduces a novel approach to variable selection that reduces the number of variables by averaging SHAP values from bootstrapped data subsets instead of using the entire dataset. This technique aims to improve computational efficiency while maintaining excellent predictive performance. To obtain classification results, we aim to train random forest, XGBoost, and C5.0 models using carefully selected variables with high interpretability. The classification accuracy of the ensemble model, generated through soft voting as the goal of high-performance model design, is compared with the individual models. The study leverages data from 1,698 Korean light industrial companies and employs bootstrapping to create distinct data groups. Logistic Regression is employed to calculate SHAP values for each data group, and their averages are computed to derive the final SHAP values. The proposed model enhances interpretability and aims to achieve superior predictive performance.

• Fire Science and Engineering
• /
• v.24 no.2
• /
• pp.67-75
• /
• 2010

The fuel moisture changes accompanying with the elapsed days after a rainfall is very important to predict the risk of forest fire and make a good use of forest fire guard. So, to investigate the conditions for the risk of forest fire, it was studied the risk of forest fire for fallen leaves level, rotten level, and soil level after more-than-5 mm-rainfall according to the different forest density of pine forests which were located in Yeong-dong region in the Spring of 2007. The result of the study showed that the around 17% of fuel moisture which was the risky level for forest fire was reached after three days of a rainfall in the coarse dense forest region and after five days in the medium or highly dense forest region. However, for the rotten level represents more than 30% of fuel moisture even after six days after the rainfall, and the lower and upper level of the soil represented a slight or almost no changes. Based on the result, the prediction model ($R^2$=0.56~0.87) for the change of fuel moisture was developed, and it was examined by applying to actual meteorological measurements in the same period of 2008. It showed a meaningful result of 1% level of distinction.

• Fire Science and Engineering
• /
• v.24 no.4
• /
• pp.18-26
• /
• 2010

The change of fuel moisture according to the passed days after a raindrop is very important to forecast risk of forest fire and to make a good use of forest fire watchmen. For that reason, in the Spring of 2007, we researched pine forest that were widespread growing in Yeongdong region to find out the condition of forest fire risk. We developed the forecast model of fuel moisture change on dead tree branches which were dropped on the ground and less than 0.6 cm, 0.6~3.0 cm, 3.0~6.0 cm, and more than 6.0 cm in diameter after more than 5.0 mm in precipitation. The result showed that the less diameter of ground fuel and small stand of pines the faster diminishing of fuel moisture, and the days of reaching to a forest fire danger fuel moisture level were represented by two (2) days for less than 0.6 cm diameter of small stand of pine and three (3) days for 0.6~3.0 cm diameter one, respectively. By those results, we developed the forecast model($R^2=0.76{\sim}0.92$) of fuel moisture change on different diameter of small stand of pine, and found that the model had statistical significant of 1% level after we applied it to the data of 2008 after the same period of raindrop by actual meteorological measurement.

• Journal of the Korea Society for Simulation
• /
• v.30 no.2
• /
• pp.49-64
• /
• 2021

M&S techniques are widely used as scientific means to systematically develop response plans to chemical and biological (CB) hazards. However, while the theoretical area of hazard dispersion modeling has achieved remarkable practical results, the operational analysis area to simulate CB hazard response plans is still in an early stage. This paper presents a model to simulate CB hazard response plans such as detection, protection, and decontamination. First, we present a possible way to display high-fidelity hazard dispersion in a combat simulation model, taking into account weather and terrain conditions. We then develop an improved vulnerability model of the combat simulation model, in order to simulate CB damage of combat simulation entities based on other casualty prediction techniques. In addition, we implement tactical behavior task models that simulate CB hazard response plans such as detection, reconnaissance, protection, and decontamination. Finally, we explore its feasibility by analyzing contamination detection effects by distributed CB detectors and decontamination effects according to the size of the {contaminated, decontamination} unit. We expect that the proposed model will be partially utilized in disaster prevention and simulation training area as well as analysis of combat effectiveness analysis of CB protection system and its operational concepts in the military area.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2011.05a
• /
• pp.382-382
• /
• 2011

최근의 극심한 기상이변으로 인하여 발생되는 유출량의 예측에 관한 사항은 치수 이수는 물론 방재의 측면에서도 역시 매우 중요한 관심사로 부각되고 있다. 강우-유출 관계는 유역의 수많은 시 공간적 변수들에 의해 영향을 받기 때문에 매우 복잡하여 예측하기 힘든 요소이며, 과거에는 추계학적 예측모형이나 확정론적 예측모형 혹은 경험적 모형 등을 사용하여 유출량을 예측하였으나 최근에는 인공신경망과 퍼지모형 그리고 유전자 알고리즘과 같은 인공지능기반의 모형들이 많이 사용되고 있다. 하지만 유출량을 예측하고자 할 때 학습자료 및 검정자료로써 사용되는 유출량은 수위-유량 관계곡선식으로부터 구하는 경우가 대부분으로 이는 이렇게 유도된 유출량의 경우 오차가 크기 때문에 그 신뢰성에 문제가 있을 것으로 판단된다. 따라서 본 논문에서는 수위를 직접 예측함으로써 이러한 오차의 문제점을 극복 하고자 한다. Neuro-Fuzzy 모형은 과거자료의 입 출력 패턴에서 정보를 추출하여 지식으로 보유하고, 이를 근거로 새로운 상황에 대한 해답을 제시하도록 하는 인공지능분야의 학습기법으로 인간이 과거의 경험과 훈련으로 지식을 축적하듯이 시스템의 입 출력에 의하여 소속함수를 최적화함으로서 모형의 구조를 스스로 조직화한다. 따라서 수학적 알고리즘의 적용이 어려운 강우와 유출관계를 하천유역이라는 시스템에서 발생된 신호체계의 입 출력패턴으로 간주하고 인간의 사고과정을 근거로 추론과정을 거쳐 수문계의 예측에 적용할 수 있을 것이다. 유전자 알고리즘은 적자생존의 생물학 원리에 바탕을 둔 최적화 기법중의 하나로 자연계의 생명체 중 환경에 잘 적응한 개체가 좀 더 많은 자손을 남길 수 있다는 자연선택 과정과 유전자의 변화를 통해서 좋은 방향으로 발전해 나간다는 자연 진화의 과정인 자연계의 유전자 메커니즘에 바탕을 둔 탐색 알고리즘이다. 즉, 자연계의 유전과 진화 메커니즘을 공학적으로 모델화함으로써 잠재적인 해의 후보들을 모아 군집을 형성한 뒤 서로간의 교배 혹은 변이를 통해서 최적 해를 찾는 계산 모델이다. 이러한 유전자 알고리즘은 전역 샘플링을 중심으로 한 수법으로 해 공간상에서 유전자의 개수만큼 복수의 탐색점을 설정할 뿐만 아니라 교배와 돌연변이 등으로 좁아지는 탐색점 바깥의 영역으로 탐색을 확장할 수 있기 때문에 지역해에 빠질 위험성이 크게 줄어든다. 따라서 예측과 패턴인식에 강한 뉴로퍼지 모형의 해 탐색방법을 유전자 알고리즘을 사용한다면 보다 정확한 해를 찾는 것이 가능할 것으로 판단된다. 따라서 본 논문에서는 선행우량 및 상류의 수위자료로부터 하류의 단시간 수위예측에 관해 연구하였으며, 이를 위해 유전자 알고리즘을 이용항여 소속함수를 최적화 시키는 형태의 Neuro-Fuzzy모형에 대하여 연구하였다.

• Journal of Korea Game Society
• /
• v.10 no.3
• /
• pp.25-35
• /
• 2010

This paper researched the possibility of applying FMEA that estimates and eliminates the failure modes into the measurement of tangible game's safety. Tangible game with actuation makes unexpected accidents for the game users. And this article tried to give risk priority number to 2 categories, game device and physical injuries using FMEA method. The result showed that TV and Hand laceration and/or bruise were revealed as the most risky factors among the others. In conclusion, it is suggested that FMEA can present integrated, quantitative and coherent measurement for the safety of tangible game.

• Korea Real Estate Review
• /
• v.28 no.3
• /
• pp.21-35
• /
• 2018

The recent performance of residential mortgages demonstrated how default risk operated separately from prepayment risk. In this study, we investigated the determinants of the borrowers' decisions pertaining to early termination through default from the mortgage performance data released by Freddie Mac, involving securitized mortgage loans from January 2011 to September 2013. We estimated a Cox-type, proportional hazard model with a single risk on fundamental factors associated with default options for individual mortgages. We proposed a mortgage default model that included two specifications of delinquency: one using a delinquency binary variable, while the other using a delinquency probability. We also compared the results obtained from two specifications with respect to goodness-of-fit proposed in the spirit of Vuong (1989) in both overlapping and nested models' cases. We found that a model with our proposed delinquency probability variable showed a statistically significant advantage compared to a benchmark model with delinquency dummy variables. We performed a default prediction power test based on the method proposed in Shumway (2001), and found a much stronger performance from the proposed model.

• Journal of the Korean Institute of Gas
• /
• v.13 no.1
• /
• pp.14-20
• /
• 2009

As the consumption of LNG has increased, the capacity and number of LNG facilities are getting bigger and bigger. Such circumstances supports the need for a dedicated risk analysis model to help review and check major issues of the safer construction and operation of LNG storage facilities systematically. Therefore this study suggests an appropriate risk analysis model that enables us to evaluate hazards of LNG storage facilities more easily and systematically, and then to use its result in siting, design and construction stages of the facilities. ill order to develop the model, lots of existing studies and domestic and foreign codes and standards were fully reviewed and a series of case studies also were carried out. The suggested model consists of 4-stage evaluations: in selecting a site, in determining a layout, in designing and constructing the facilities, and in operating them. This model also suggests the weather condition necessary for estimating the consequence of accident-scenarios, and the easy, systematic approach to the analysis of their probability. We expect that the model may help secure LNG storage facilities' inherent safety in determining their site and layout.