• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2023년도 학술발표회
• /
• pp.31-31
• /
• 2023

During December 2022, the northeast monsoon, which dominates the south and the Gulf of Thailand, had significant rainfall that impacted the lower southern region, causing flash floods, landslides, blustery winds, and the river exceeding its bank. The Golok River, located in Narathiwat, divides the border between Thailand and Malaysia was also affected by rainfall. In flood management, instruments for measuring precipitation and water level have become important for assessing and forecasting the trend of situations and areas of risk. However, such regions are international borders, so the installed measuring telemetry system cannot measure the rainfall and water level of the entire area. This study aims to predict 72 hours of water level and evaluate the situation as information to support the government in making water management decisions, publicizing them to relevant agencies, and warning citizens during crisis events. This research is applied to machine learning (ML) for water level prediction of the Golok River, Lan Tu Bridge area, Sungai Golok Subdistrict, Su-ngai Golok District, Narathiwat Province, which is one of the major monitored rivers. The eXtreme Gradient Boosting (XGBoost) algorithm, a tree-based ensemble machine learning algorithm, was exploited to predict hourly water levels through the R programming language. Model training and testing were carried out utilizing observed hourly rainfall from the STH010 station and hourly water level data from the X.119A station between 2020 and 2022 as main prediction inputs. Furthermore, this model applies hourly spatial rainfall forecasting data from Weather Research and Forecasting and Regional Ocean Model System models (WRF-ROMs) provided by Hydro-Informatics Institute (HII) as input, allowing the model to predict the hourly water level in the Golok River. The evaluation of the predicted performances using the statistical performance metrics, delivering an R-square of 0.96 can validate the results as robust forecasting outcomes. The result shows that the predicted water level at the X.119A telemetry station (Golok River) is in a steady decline, which relates to the input data of predicted 72-hour rainfall from WRF-ROMs having decreased. In short, the relationship between input and result can be used to evaluate flood situations. Here, the data is contributed to the Operational support to the Special Water Resources Management Operation Center in Southern Thailand for flood preparedness and response to make intelligent decisions on water management during crisis occurrences, as well as to be prepared and prevent loss and harm to citizens.

• 한국철도학회논문집
• /
• 제13권1호
• /
• pp.71-77
• /
• 2010

대부분의 선진국에서 교량의 유지보수 및 보강(Maintenance Repair & Rehabilitation-MR&R)으로 인한 비용은 해마다 증가하고 있다. 전산화된 교량유지관리 및 의사결정시스템(Bridge Management System-BMS)은 가능한 최저의 생애주기비용(Life Cycle Cost - LCC)에 최적의 안정성를 확보하기 위해 개발되었다. 본 논문에서는 제한된 현존하는 교량진단기록을 이용하여 현존하지 않는 과거의 교량상태등급 데이타를 생성하기 위해 Backward Prediction Model(BPM)이라 불리는 인공신경망(Artificial Neural Network-ANN)에 기초한 예측모델을 제시한다. 제안된 BPM은 한정된 교량 정기점검기록으로부터 현존하는 교량진단기록과 연관성을 확립하기 위해 교통량과 인구, 그리고 기후 등과 같은 비구조적 요소를 이용하며, 제한된 교량진단기록과 비구조적 요소 사이에 맺어진 연관성을 통해 현존하지 않는 과거의 교량상태등급 데이타를 생성할 수 있다. BPM의 신뢰도를 측정하기 위하여 Maryland DOT로 부터 얻어진 National Bridge Inventory(NBI)와 BMS 교량진단자료를 이용하였다. 이중 NBI자료를 이용한 Backward comparison 에 있어서 실제 NBI기록과 BPM으로 생성된 교량상태등급과의 차이(상판: 6.68%, 상부구조부: 6.61%, 하부구조부: 7.52%)는 BPM으로 생성된 결과의 높은 신뢰도를 보여준다. 이 연구의 결과는 제한된 정기점검 기록으로 야기되는 BMS의 장기 교량손상 예측에 관련된 사용상의 문제를 최소화하고 전반적인 BMS 결과의 신뢰도를 높이는데 기여 할 수 있다.

• 한국ITS학회 논문지
• /
• 제12권4호
• /
• pp.44-55
• /
• 2013

지능형교통체계(ITS: Intelligent Transportation System)의 발전은 과거에 비해 보다 신뢰성 있고 폭넓은 교통자료 및 기상자료 등의 취득을 가능하도록 하였다. 이러한 첨단 시스템의 발전에 따라 수집된 자료를 이용하여 교통상황과 기상상황에 대한 다양한 연구가 활발히 진행되고 있다. 본 연구에서는 도로 기상정보 시스템(RWIS: Road Weather Information System)자료와 검지기 자료를 이용하여 강우량에 따른 속도 감소 패턴을 분석하고, 강우량에 따른 속도감소량 산출 결과를 통해 강우수준을 분류하는 기준을 제시하였다. 인공신경망을 이용하여 강우수준별 통행속도를 예측하였으며, 예측 결과를 비교하여 강우수준별 통행속도 예측 특성을 분석하였다. 분석결과, 강우수준 분류 기준은 0.4mm/5min, 0.8mm/5min으로 나타났으며, 강우수준별 속도와 교통량에 대한 분산분석 결과 강우수준별로 차이를 보이는 것으로 나타났다. 인공신경망을 통한 5분 단위의 통행속도 예측결과, 비강우인 경우에는 과거 5개 자료, 즉, 25분 동안의 속도자료를 사용하여 분석하는 것이 예측력이 높게 나타났으며, 강우가 발생하는 경우에는 과거 2~3개 자료, 즉, 10~15분 동안의 속도자료를 사용하는 것이 예측력이 높게 나타났다. 본 연구에서는 기상조건에 관계없이 신뢰성 있는 교통정보를 제공하기 위한 통행시간 예측 방법론을 제시함으로써 통행시간 정보 등의 교통정보 제공 시 보다 정확한 정보를 제공하여 교통상황 예측정보의 신뢰도 향상 및 교통상황 예측정보의 활용도를 증대시킬 수 있을 것으로 기대된다.

• Clinical and Experimental Pediatrics
• /
• 제59권3호
• /
• pp.139-144
• /
• 2016

Purpose: Acute pyelonephritis (APN) is a serious bacterial infection that can cause renal scarring in children. Early identification of APN is critical to improve treatment outcomes. The neutrophil-lymphocyte ratio (NLR) is a prognostic marker of many diseases, but it has not yet been established in urinary tract infection (UTI). The aim of this study was to determine whether NLR is a useful marker to predict APN or vesicoureteral reflux (VUR). Methods: We retrospectively evaluated 298 pediatric patients ($age{\leq}36months$) with febrile UTI from January 2010 to December 2014. Conventional infection markers (white blood cell [WBC] count, erythrocyte sedimentation rate [ESR], C-reactive protein [CRP]), and NLR were measured. Results: WBC, CRP, ESR, and NLR were higher in APN than in lower UTI (P<0.001). Multiple logistic regression analyses showed that NLR was a predictive factor for positive dimercaptosuccinic acid (DMSA) defects (P<0.001). The area under the receiver operating characteristic (ROC) curve was high for NLR (P<0.001) as well as CRP (P<0.001) for prediction of DMSA defects. NLR showed the highest area under the ROC curve for diagnosis of VUR (P<0.001). Conclusion: NLR can be used as a diagnostic marker of APN with DMSA defect, showing better results than those of conventional markers for VUR prediction.

• Computers and Concrete
• /
• 제22권2호
• /
• pp.249-259
• /
• 2018

This paper presents Artificial Neural Network (ANN) models for evaluating bond strength of deformed, plain and cold formed bars in low strength concrete. The ANN models were implemented using the experimental database developed by conducting experiments in three different universities on total of 138 pullout and 108 splitting specimens under monotonic loading. The key parameters examined in the experiments are low strength concrete, bar development length, concrete cover, rebar type (deformed, cold-formed, plain) and diameter. These deficient parameters are typically found in non-engineered reinforced concrete structures of developing countries. To develop ANN bond model for each bar type, four inputs (the low strength concrete, development length, concrete cover and bar diameter) are used for training the neurons in the network. Multi-Layer-Perceptron was trained according to a back-propagation algorithm. The ANN bond model for deformed bar consists of a single hidden layer and the 9 neurons. For Tor bar and plain bars the ANN models consist of 5 and 6 neurons and a single hidden layer, respectively. The developed ANN models are capable of predicting bond strength for both pull and splitting bond failure modes. The developed ANN models have higher coefficient of determination in training, validation and testing with good prediction and generalization capacity. The comparison of experimental bond strength values with the outcomes of ANN models showed good agreement. Moreover, the ANN model predictions by varying different parameters are also presented for all bar types.

• Korean Journal of Ophthalmology
• /
• 제32권6호
• /
• pp.497-505
• /
• 2018

Purpose: To evaluate and compare published methods of calculating intraocular lens (IOL) power following myopic laser refractive surgery. Methods: We performed a retrospective review of the medical records of 69 patients (69 eyes) who had undergone myopic laser refractive surgery previously and subsequently underwent cataract surgery at Samsung Medical Center in Seoul, South Korea from January 2010 to June 2016. None of the patients had pre-refractive surgery biometric data available. The Haigis-L, Shammas, Barrett True-K (no history), Wang-Koch-Maloney, Scheimpflug total corneal refractive power (TCRP) 3 and 4 mm (SRK-T and Haigis), Scheimpflug true net power, and Scheimpflug true refractive power (TRP) 3 mm, 4 mm, and 5 mm (SRK-T and Haigis) methods were employed. IOL power required for target refraction was back-calculated using stable post-cataract surgery manifest refraction, and implanted IOL power and formula accuracy were subsequently compared among calculation methods. Results: Haigis-L, Shammas, Barrett True-K (no history), Wang-Koch-Maloney, Scheimpflug TCRP 4 mm (Haigis), Scheimpflug true net power 4 mm (Haigis), and Scheimpflug TRP 4 mm (Haigis) formulae showed high predictability, with mean arithmetic prediction errors and standard deviations of $-0.25{\pm}0.59$, $-0.05{\pm}1.19$, $0.00{\pm}0.88$, $-0.26{\pm}1.17$, $0.00{\pm}1.09$, $-0.71{\pm}1.20$, and $0.03{\pm}1.25$ diopters, respectively. Conclusions: Visual outcomes within 1.0 diopter of target refraction were achieved in 85% of eyes using the calculation methods listed above. Haigis-L, Barrett True-K (no history), and Scheimpflug TCRP 4 mm (Haigis) and TRP 4 mm (Haigis) methods showed comparably low prediction errors, despite the absence of historical patient information.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1997년도 추계학술대회발표논문집; 홍익대학교, 서울; 1 Nov. 1997
• /
• pp.105-108
• /
• 1997

Previous studies in stock market predictions using artificial intelligence techniques such as artificial neural networks and case-based reasoning, have focused mainly on spot market prediction. Korea launched trading in index futures market (KOSPI 200) on May 3, 1996, then more people became attracted to this market. Thus, this research intends to predict the daily up/down fluctuant direction of the price for KOSPI 200 index futures to meet this recent surge of interest. The forecasting methodologies employed in this research are the integration of genetic algorithm and artificial neural network (GAANN) and the integration of genetic algorithm and case-based reasoning (GACBR). Genetic algorithm was mainly used to select relevant input variables. This study adopts the categorical data preprocessing based on expert's knowledge as well as traditional data preprocessing. The experimental results of each forecasting method with each data preprocessing method are compared and statistically tested. Artificial neural network and case-based reasoning methods with best performance are integrated. Out-of-the Model Integration and In-Model Integration are presented as the integration methodology. The research outcomes are as follows; First, genetic algorithms are useful and effective method to select input variables for Al techniques. Second, the results of the experiment with categorical data preprocessing significantly outperform that with traditional data preprocessing in forecasting up/down fluctuant direction of index futures price. Third, the integration of genetic algorithm and case-based reasoning (GACBR) outperforms the integration of genetic algorithm and artificial neural network (GAANN). Forth, the integration of genetic algorithm, case-based reasoning and artificial neural network (GAANN-GACBR, GACBRNN and GANNCBR) provide worse results than GACBR.

• International journal of advanced smart convergence
• /
• 제10권2호
• /
• pp.145-150
• /
• 2021

Patients with Pressure injuries (PIs) may have pain and discomfort, which results in poorer patient outcomes and additional cost for treatment. This study was a part of larger research project that aimed at prediction modeling using a big data. The purpose of this study were to describe the characteristics of patients with PI in critical care; and to explore comorbidity and diagnostic and interventive procedures that have been done for patients in critical care. This is a secondary data analysis. Data were retrieved from a large clinical database, MIMIC-III Clinical database. The number of unique patients with PI was 2,286 in total. Approximately 60% were male and 68.4% were White. Among the patients, 9.9% were dead. In term of discharge disposition, 56.2% (33.9% Home, 22.3% Home Health Care) where as 32.3% were transferred to another institutions. The rest of them were hospice (0.8%), left against medical advice (0.7%), and others (0.2%). The top three most frequently co-existing kinds of diseases were Hypertension, not otherwise specified (NOS), congestive heart failure NOS, and Acute kidney failure NOS. The number of patients with PI who have one or more procedures was 2,169 (94.9%). The number of unique procedures was 981. The top three most frequent procedures were 'Venous catheterization, not elsewhere classified,' and 'Enteral infusion of concentrated nutritional substances.' Patient with a greater number of comorbid conditions were likely to have longer length of ICU stay (r=.452, p<.001). In addition, patient with a greater number of procedures that were performed during the admission were strongly tend to stay longer in hospital (r=.729, p<.001). Therefore, prospective studies focusing on comorbidity; and diagnostic and preventive procedures are needed in the prediction modeling of pressure injury development in ICU patients.

• 해양환경안전학회지
• /
• 제30권1호
• /
• pp.82-88
• /
• 2024

현대 해양 산업은 기술적 발전을 통해 신속한 발전을 이루고 있다. 이러한 발전을 주도하는 주요 기술 중 하나는 데이터 처리 기술이며, 이 중 자연어 처리 기법은 사람의 언어를 기계가 이해하고 처리할 수 있도록 하는 기술이다. 본 연구는 자연어 처리 기법을 통해 해양안전심판원의 재결서를 분석하여 이미 재결이 이루어진 선박 충돌사고의 원인 제공 비율을 학습한 후, 새로운 재결서를 입력하면 원인 제공 비율을 예측하는 모델을 개발하고자 하였다. 이 모델은 사고 당시 적용되는 항법과 원인 제공 비율에 영향을 주는 핵심 키워드의 가중치를 이용하여 사고의 원인 제공 비율을 계산하는 방식으로 구성하였다. 이 연구는 이러한 방식을 통해 제작한 모델의 정확도를 분석하고, 모델의 실무 적용 가능성을 검토함과 동시에 충돌사고 재발 방지 및 해양사고 당사자들의 분쟁 해결에 기여할 것으로 기대한다.

• Korean Journal of Radiology
• /
• 제22권2호
• /
• pp.213-224
• /
• 2021

Objective: Clinical outcomes of patients who undergo transarterial chemoembolization (TACE) for single small hepatocellular carcinoma (HCC) are not consistent, and may differ based on certain imaging findings. This retrospective study was aimed at determining the efficacy of pre-TACE CT or MR imaging findings in predicting survival outcomes in patients with small HCC upon being treated with TACE. Besides, the study proposed to build a risk prediction model for these patients. Materials and Methods: Altogether, 750 patients with functionally good hepatic reserve who received TACE as the first-line treatment for single small HCC between 2004 and 2014 were included in the study. These patients were randomly assigned into training (n = 525) and validation (n = 225) sets. Results: According to the results of a multivariable Cox analysis, three pre-TACE imaging findings (tumor margin, tumor location, enhancement pattern) and two clinical factors (age, serum albumin level) were selected and scored to create predictive models for overall, local tumor progression (LTP)-free, and progression-free survival in the training set. The median overall survival time in the validation set were 137.5 months, 76.1 months, and 44.0 months for low-, intermediate-, and high-risk groups, respectively (p < 0.001). Time-dependent receiver operating characteristic curves of the predictive models for overall, LTP-free, and progression-free survival applied to the validation cohort showed acceptable areas under the curve values (0.734, 0.802, and 0.775 for overall survival; 0.738, 0.789, and 0.791 for LTP-free survival; and 0.671, 0.733, and 0.694 for progression-free survival at 3, 5, and 10 years, respectively). Conclusion: Pre-TACE CT or MR imaging findings could predict survival outcomes in patients with small HCC upon treatment with TACE. Our predictive models including three imaging predictors could be helpful in prognostication, identification, and selection of suitable candidates for TACE in patients with single small HCC.