• 한국재난정보학회 논문집
• /
• 제17권2호
• /
• pp.245-253
• /
• 2021

연구목적: 본 연구는 화재발생 건축물 정보, 신고자 취득 정보 등 초기 정보를 활용하여 화재현장의 위험도를 예측하여, 재난 발생 초기에 효과적인 소방자원 동원 및 적절한 대응을 위한 피해최소화 전략 수립을 지원하는 위험도 예측 모델을 개발하고자 한다. 연구방법: 화재 통계 데이터 상에서 화재의 피해규모와 관련된 변수 규명을 위해 머신러닝 알고리즘을 이용한 변수간 상관성 분석을 실시하여 예측 가능성을 검토하고, 데이터 표준화 및 이산화 등의 전처리를 통해 학습 데이터 셋을 구축하였다. 이를 활용하여 예측 정확도가 높은 것으로 평가 받고 있는 복수의 머신러닝 알고리즘을 테스트하여 가장 정확도가 높은 알고리즘을 적용한 위험도 예측 모델을 개발하였다. 연구결과: 머신러닝 알고리즘 성능 테스트 결과 랜덤포레스트 알고리즘의 정확도가 가장 높게 나왔으며, 위험도 등급에 대해서는 중간치에 대한 정확성이 상대적으로 높은 것으로 확인되었다. 결론: 화재 통계 상 피해규모 데이터의 편향성에 의해 예측모델 정확도가 제한적으로 나타났으며, 예측 모델 성능 개선을 위해 데이터 정합성 및 결손치 보완 등을 통한 데이터 정제가 필요하다.

• 한국건축시공학회:학술대회논문집
• /
• 한국건축시공학회 2023년도 가을학술발표대회논문집
• /
• pp.251-252
• /
• 2023

Compared to other industries the construction industry experiences more casualties and property damage due to safety accidents. One of the reasons is the increasing number of foreign workers. For this reason, past studies have found that foreign workers at construction sites are more exposed to safety accidents than non-foreign workers. Nevertheless the proportion of foreign workers involved in safety accidents at construction sites is increasing, and there has been a lack of research to predict the risk of safety accidents at construction sites. Additionally, realistic safety management is lacking due to a lack of safety accident risk prediction research. Therefore, in this study, we would like to propose basic research that proposes an AI-based safety accident prediction model framework for predicting safety accidents of foreign workers at construction sites. The framework and results of this study will contribute to reducing and preventing the risk of safety accidents for foreign workers through risk prediction for safety management of foreign workers at construction sites.

• 한국화재소방학회논문지
• /
• 제33권4호
• /
• pp.50-58
• /
• 2019

구획실 내 가연물들의 화재거동에 대한 B-RISK의 예측성능을 Fire dynamics simulator (FDS)와 연계하여 검토하였다. 먼저 열발생률(Heat release rate, HRR)에 대한 B-RISK의 예측성능을 검토하기 위해 가연물 한 세트의 실험에서 측정된 HRR 값과 디자인 화재곡선을 B-RISK의 입력조건으로 사용하여 가연물 두 세트에 대한 HRR 곡선을 계산하고 실험에서 측정된 가연물 두 세트의 HRR 값과 비교하였다. B-RISK 결과와 실험결과를 비교하여 B-RISK가 화재성장률에 대한 예측은 어렵지만 최대 HRR 값과 총 열발생량에 대해서는 충분히 예측할 수 있음을 확인하였다. 그리고 B-RISK 계산을 통해 예측된 HRR 값을 FDS의 입력조건으로 사용하여 계산된 결과와 실험결과를 비교하여 B-RISK 계산을 통해 예측된 HRR 값의 화재거동에 대해 검토하였다. 실험에서 측정된 온도 및 화학종 농도 결과와 비교하여 화재성장구간에 대해 차이가 있는 것을 확인하였지만 예측된 HRR 값에서 Percentile이 약 70%인 HRR 값을 사용하더라도 충분히 전체적인 화재거동을 예측할 수 있음을 확인하였다.

• Korean Circulation Journal
• /
• 제53권9호
• /
• pp.581-593
• /
• 2023

In the catheterization laboratory, the measurement of physiological indexes can help identify functionally significant lesions and has become one of the standard methods to guide treatment decision-making. Plaque vulnerability refers to a coronary plaque susceptible to rupture, enabling risk prediction before coronary events, and it can be detected by defining a certain type of plaque morphology on coronary imaging modalities. Although coronary physiology and plaque vulnerability have been considered different attributes of coronary artery disease, the underlying pathophysiological basis and clinical data indicate a strong correlation between coronary hemodynamic properties and vulnerable plaque. In prediction of coronary events, emerging data have suggested independent and additional implications of a physiology-based approach to a plaque-based approach. This review covers the fundamental interplay between coronary physiology and plaque morphology during disease progression with clinical data supporting this relationship and examines the clinical relevance of physiological indexes in prediction of clinical outcomes and therapeutic decision-making along with plaque vulnerability.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제18권1호
• /
• pp.30-45
• /
• 2024

This study introduces an effective method for predicting individual local tax delinquencies using prevalent machine learning and deep learning algorithms. The evaluation of credit risk holds great significance in the financial realm, impacting both companies and individuals. While credit risk prediction has been explored using statistical and machine learning techniques, their application to tax arrears prediction remains underexplored. We forecast individual local tax defaults in Republic of Korea using machine and deep learning algorithms, including convolutional neural networks (CNN), long short-term memory (LSTM), and sequence-to-sequence (seq2seq). Our model incorporates diverse credit and public information like loan history, delinquency records, credit card usage, and public taxation data, offering richer insights than prior studies. The results highlight the superior predictive accuracy of the CNN model. Anticipating local tax arrears more effectively could lead to efficient allocation of administrative resources. By leveraging advanced machine learning, this research offers a promising avenue for refining tax collection strategies and resource management.

• Genomics & Informatics
• /
• 제10권3호
• /
• pp.175-183
• /
• 2012

Colorectal cancer (CRC) is among the leading causes of cancer deaths and can be caused by environmental factors as well as genetic factors. Therefore, we developed a prediction model of CRC using genetic risk scores (GRS) and evaluated the effects of conventional risk factors, including family history of CRC, in combination with GRS on the risk of CRC in Koreans. This study included 187 cases (men, 133; women, 54) and 976 controls (men, 554; women, 422). GRS were calculated with most significantly associated single-nucleotide polymorphism with CRC through a genomewide association study. The area under the curve (AUC) increased by 0.5% to 5.2% when either counted or weighted GRS was added to a prediction model consisting of age alone (AUC 0.687 for men, 0.598 for women) or age and family history of CRC (AUC 0.692 for men, 0.603 for women) for both men and women. Furthermore, the risk of CRC significantly increased for individuals with a family history of CRC in the highest quartile of GRS when compared to subjects without a family history of CRC in the lowest quartile of GRS (counted GRS odds ratio [OR], 47.9; 95% confidence interval [CI], 4.9 to 471.8 for men; OR, 22.3; 95% CI, 1.4 to 344.2 for women) (weighted GRS OR, 35.9; 95% CI, 5.9 to 218.2 for men; OR, 18.1, 95% CI, 3.7 to 88.1 for women). Our findings suggest that in Koreans, especially in Korean men, GRS improve the prediction of CRC when considered in conjunction with age and family history of CRC.

• Korean Circulation Journal
• /
• 제53권9호
• /
• pp.621-631
• /
• 2023

Background and Objectives: The morphology-voltage-P-wave duration (MVP) electrocardiography (ECG) risk score is a newly defined scoring system that has recently been used for atrial fibrillation (AF) prediction. The aim of this study was to evaluate the ability of the MVP ECG risk score to predict AF in patients with an implantable cardioverter defibrillator (ICD) and heart failure with reduced ejection fraction in long-term follow-up. Methods: The study used a single-center, and retrospective design. The study included 328 patients who underwent ICD implantation in our hospital between January 2010 and April 2021, diagnosed with heart failure. The patients were divided into low, intermediate and high-risk categories according to the MVP ECG risk scores. The long-term development of atrial fibrillation was compared among these 3 groups. Results: The low-risk group included 191 patients, the intermediate-risk group 114 patients, and the high-risk group 23 patients. The long-term AF development rate was 12.0% in the low-risk group, 21.9% in the intermediate risk group, and 78.3% in the high-risk group. Patients in the high-risk group were found to have 5.2 times higher rates of long-term AF occurrence compared to low-risk group. Conclusions: The MVP ECG risk score, which is an inexpensive, simple and easily accessible tool, was found to be a significant predictor of the development of AF in the long-term follow-up of patients with an ICD with heart failure with reduced ejection fraction. This risk score may be used to identify patients who require close follow-up for development and management of AF.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2009년도 춘계학술대회 논문집
• /
• pp.941-942
• /
• 2009

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2009년도 춘계학술대회 논문집
• /
• pp.53-54
• /
• 2009

• 한국재난정보학회 논문집
• /
• 제19권2호
• /
• pp.334-343
• /
• 2023

연구목적: 본 연구에서는 지반공동탐사로 발견된 공동자료를 지하시설물과의 원인별 상관관계로 분석하고, AI 알고리즘 기반으로 지반침하 예측지도를 검증하여 시민에게 안전한 도로 환경을 제공하고자한다. 연구방법: 위험도 평가 관련 데이터조사와 빅데이터 수집, AI분석을 위한 데이터 전처리, 그리고 AI 알고리즘을 이용하여 지반침하 위험도 예측지도 검증 등 3가지 단계로 연구를 수행하였다. 연구결과:작성한 지반침하 위험 예측지도를 분석하여 부산시 부산진구와 사하구에 대해 긴급, 우선, 일반 3단계의 공동관리 위험등급 분포를 확인 할 수 있었다. 또한, 지반침하 위험 등급 예측 값을 도로노선의 구간별로 정리하여 긴급 등급이 포함된 도로가 부산진구는 총 61개구간 중 3개소, 사하구는 총 68개구간 중 7개소임을 확인하였으며 각 도로노선별 지반침하 위험 예측 순위를 파악하였다. 결론: 도출된 지반침하 위험 예측지도를 바탕으로 효율적으로 탐사구간을 설정하여 우선 조사, 선제 조치함으로써 시민들의 불안을 해소하고 효율적인 도로유지관리 및 보수, 제도의 개선 등의 부수적인 효과를 얻을 수 있다.