유전자 알고리즘을 적용한 댐 운영의 최적화 방만에 관한 연구

신홍준,허준행 한국수자원학회 2001 한국수자원학회 학술대회 Vol.2001 No.1

Paradoxically negative response to bronchodilator in patients with chronic obstructive pulmonary disease : The KOrea COpd Subgroup Study Team Cohort

신홍준,김민석,박하영,고보건,김태옥,박철규,오인재,김유일,김영철,임성철 대한결핵 및 호흡기학회 2018 대한결핵 및 호흡기학회 추계학술대회 초록집 Vol.126 No.0

Aim: We assessed the clinical characteristics and outcomes in patients who were paradoxically decreased FEV1 after bronchodilator administration from KOCOSS cohort. Methods: Paradoxically negative response (PR) was defined as reduction in post-bronchodilator FEV1 compared to pre-bronchodilator FEV1. Negative bronchodilator response (negative BDR) was defined as at least a 12% and 200 mL reduction in FEV1 or FVC, or both, after administration of bronchodilator. Results: Total 1,991 patients of the cohort were recruited from 2012 to 2017. PR was noted in 369 (18.5%) of 1,991 patients. Compared to the non-PR group, PR group had higher pre-bronchodilator FEV1, and C-reactive protein, and lower 6-minute walk distance (6MWD). PR was independently associated with worse dyspnea (modified Medical Research Council dyspnea scale, COPD assessment test), lower 6MWD, and poor quality of life (St. George’s Respiratory Questionnaire). Pre-bronchodilator FEV1 (L) had better area under curve (AUC) for predicting 1-year severe acute exacerbation than post-bronchodilator FEV1 (L) in the PR group (0.690 vs. 0.663). Negative BDR was noted in 57 (2.9%) of 1,991 patients. Pre-bronchodilator FEV1 (L) had better AUC for predicting 1-year severe acute exacerbation than post-bronchodilator FEV1 (L) in the negative BDR group (0.788 vs. 0.752). Both PR and negative BDR were not associated with 1-year severe exacerbation. Conclusion: Pre-bronchodilator FEV1 (L) could be better helpful predicting 1-year acute exacerbation in patients with paradoxically decreased FEV1 after bronchodilator administration.

한수원 수력 댐 강우예측 AI 경진대회 개최 결과

신홍준,윤성심,Sin, Hong-Jun,Yun, Seong-Sim 한국수자원학회 2021 물과 미래(한국수자원학회지) Vol.54 No.1

시간 연속성을 고려한 딥러닝 기반 레이더 강우예측

신홍준,윤성심,최재민 한국수자원학회 2021 한국수자원학회논문집 Vol.54 No.5

In this study, we tried to improve the performance of the existing U-net-based deep learning rainfall prediction model, which can weaken the meaning of time series order. For this, ConvLSTM2D U-Net structure model considering temporal consistency of data was applied, and we evaluated accuracy of the ConvLSTM2D U-Net model using a RainNet model and an extrapolation-based advection model. In addition, we tried to improve the uncertainty in the model training process by performing learning not only with a single model but also with 10 ensemble models. The trained neural network rainfall prediction model was optimized to generate 10-minute advance prediction data using four consecutive data of the past 30 minutes from the present. The results of deep learning rainfall prediction models are difficult to identify schematically distinct differences, but with ConvLSTM2D U-Net, the magnitude of the prediction error is the smallest and the location of rainfall is relatively accurate. In particular, the ensemble ConvLSTM2D U-Net showed high CSI, low MAE, and a narrow error range, and predicted rainfall more accurately and stable prediction performance than other models. However, the prediction performance for a specific point was very low compared to the prediction performance for the entire area, and the deep learning rainfall prediction model also had limitations. Through this study, it was confirmed that the ConvLSTM2D U-Net neural network structure to account for the change of time could increase the prediction accuracy, but there is still a limitation of the convolution deep neural network model due to spatial smoothing in the strong rainfall region or detailed rainfall prediction. 본 연구에서는 시계열 순서의 의미가 희석될 수 있는 기존의 U-net 기반 딥러닝 강우예측 모델의 성능을 개선하고자 하였다. 이를 위해서 데이터의 연속성을 고려한 ConvLSTM2D U-Net 신경망 구조를 갖는 모델을 적용하고, RainNet 모델 및 외삽 기반의 이류모델을 이용하여 예측정확도 개선 정도를 평가하였다. 또한 신경망 기반 모델 학습과정에서의 불확실성을 개선하기 위해 단일 모델뿐만 아니라 10개의 앙상블 모델로 학습을 수행하였다. 학습된 신경망 강우예측모델은 현재를 기준으로 과거 30분 전까지의 연속된 4개의 자료를 이용하여 10분 선행 예측자료를 생성하는데 최적화되었다. 최적화된 딥러닝 강우예측모델을 이용하여 강우예측을 수행한 결과, ConvLSTM2D U-Net을 사용하였을 때 예측 오차의 크기가 가장 작고, 강우 이동 위치를 상대적으로 정확히 구현하였다. 특히, 앙상블 ConvLSTM2D U-Net이 타 예측모델에 비해 높은 CSI와 낮은 MAE를 보이며, 상대적으로 정확하게 강우를 예측하였으며, 좁은 오차범위로 안정적인 예측성능을 보여주었다. 다만, 특정 지점만을 대상으로 한 예측성능은 전체 강우 영역에 대한 예측성능에 비해 낮게 나타나, 상세한 영역의 강우예측에 대한 딥러닝 강우예측모델의 한계도 확인하였다. 본 연구를 통해 시간의 변화를 고려하기 위한 ConvLSTM2D U-Net 신경망 구조가 예측정확도를 높일 수 있었으나, 여전히 강한 강우영역이나 상세한 강우예측에는 공간평활로 인한 합성곱 신경망 모델의 한계가 있음을 확인하였다.

강우자료의 비정상성을 고려한 재현기간 변화에 관한 연구

신홍준,안현준,허준행 한국수자원학회 2014 한국수자원학회논문집 Vol.47 No.5

본 연구에서는 초과확률 또는 비초과확률이 시간에 따라 변화한다는 비정상성을 가정하여 재현기간 산정에 대한 연구를 수행하였다. 비정상성을 고려한 2가지 재현기간 산정 방법에 대해 검토하고 비정상성 Gumbel 모형을 이용한 빈도해석을 수행하여 초과확률 및 비초과확률을 구한 뒤 비정상성을 고려한 재현기간 정의에 따른 우리나라 재현기간의 변화에 대해서 살펴보았다. 적용 대상으로는 자료기간 30년 이상을 보유하면서 일 강우 자료의 경향성이 나타나는 서귀포, 인제, 제천, 구미, 문경, 거창 등 6개 지점을 선정하였다. 적용결과 비정상성을 고려한 재현기간 산정 시 기존의 재현기간 산정 방법과는 재현기간이 다르게 산정됨을 알 수 있었고, 재현기간이 커질수록 정상성 가정하의 재현기간과 비정상성 가정하의 재현기간 값의 차이가 더 커지는 것으로 나타났다. 또한 비정상성을 고려한 재현기간의 2가지 정의 중 기대 대기시간(expected waiting time) 정의에 의한 방법이 기대 초과사상 수(expected number of exceedance event) 정의에 의한 방법 보다 작은 재현기간이 산정 되었다. This research focuses on the changes of return period for nonstationary rainfall data in which exceedance or nonexceedance probability varies depending on time. We examined two definitions of return period under nonstationarity and also performed nonstationary frequency analysis using the nonstationary Gumbel model to investigate variations of return period in Korea. Seogwipo, Inje, Jecheon, Gumi, Mungyeong, and Geochang were selected as subject sites of application. These sites have a trend in rainfall data as well as having more than 30 years data. As the results of application, the return periods considering nonstationarity are different with those considering stationarity. The differences of return periods between nonstationarity and stationarity increase as growing return period increases. In addition, the return period using the expected waiting time method shows lower value than that using the expected number of event method.

Treatment of Drug Susceptible Pulmonary Tuberculosis

신홍준,권용수 대한결핵및호흡기학회 2015 Tuberculosis and Respiratory Diseases Vol.78 No.3

Tuberculosis (TB) remains a major global health problem, and the incidence of TB cases has not significantly decreased over the past decade in Korea. The standard short course regimen is highly effective against TB, but requires multiple TB-specific drugs and a long treatment duration. Recent studies using late-generation fluoroquinolones and/or high-dose rifapentine-containing regimens to shorten the duration of TB treatment showed negative results. Extending the treatment duration may be considered in patients with cavitation on the initial chest radiograph and positivity in sputum culture at 2 months of treatment for preventing TB relapse. Current evidence does not support the use of fixed-dose combinations compared to separate drugs for the purpose of improving treatment outcomes. All patients receiving TB treatment should be monitored regularly for response to therapy, facilitation of treatment completion, and management of adverse drug reactions. Mild adverse effects can be managed with symptomatic therapy and changing the timing of the drug administration, but severe adverse effects require a discontinuation of the offending drugs.

Monte Carlo 모의를 이용한 지점 간 확률가중모멘트의 교차상관관계

신홍준,정영훈,허준행 한국수자원학회 2009 한국수자원학회논문집 Vol.42 No.3

본 연구에서는 지점 간 확률가중모멘트의 교차상관관계를 구하기 위해 Monte Carlo 모의를 이용하여 이를 지점간 표본자료의 교차상관성에 대한 관계식으로 확장하여 근사값을 구하고자 하였다. 모의실험 결과 각각의 확률가중 모멘트간 교차상관계수는 지점 간 표본자료의 교차상관계수와 자료크기가 동일하고 동시간 자료일 경우 기울기 1인 선형관계를 보이며, 자료크기에 따른 동시간 자료의 비율이 작아질수록 선형적인 관계는 점점 약해지게 된다. 따라서 자료크기에 따 In this study, cross correlations among sample data at each site are calculated to obtain the asymptotic cross correlations among probability weighted moments at each site using Monte Carlo simulation. As a result, the relations between the asymptotic cro

강우 지역빈도해석의 적용성 연구

신홍준,남우성,허준행,김경덕 한국수자원학회 2005 한국수자원학회 학술대회 Vol.2005 No.1

Generalized Logisitc 분포형을 이용한 위험도에 관한 불확실성 해석

신홍준,정영훈,김수영,허준행 대한토목학회 2006 대한토목학회 학술대회 Vol.2006 No.10

When is LABA/LAMA Better than LAMA in GOLD Group B or D Patients for Reducing Acute Exacerbations of COPD?

신홍준,김유일,김유림,이창열,라승원,문지용,정기석,유광하,신경철,임성철,KOCOSS Cohort Study Group 전남대학교 의과학연구소 2023 전남의대학술지 Vol.59 No.3

Long-acting b2-agonist (LABA)/long-acting muscarinic-antagonist (LAMA) dual therapy has been found to be more effective than LAMA monotherapy in the treatment of chronic obstructive pulmonary disease (COPD). However, among patients with group B or D COPD, the characteristics of patients for whom LABA/LAMA dual therapy is superior to LAMA monotherapy in minimizing acute exacerbations remain unknown. With data from a prospective COPD cohort, subgroup analyses were conducted to determine whether LABA/LAMA dual therapy was superior to LAMA monotherapy in reducing the rate of acute exacerbations in group B and D COPD patients. Group B and D COPD patients taking LAMA or LABA/LAMA were enrolled according to the 2022 Global initiative for Chronic Obstructive Pulmonary Disease guidelines. A total of 737 patients were included in this study: 600 with group B COPD and 137 with group D COPD. Compared with patients taking LAMA monotherapy, those taking LABA/ LAMA had a significantly lower incidence of acute exacerbations over 1 year. In the subgroup of patients ≥70 years old, there was a significantly lower risk of severe COPD exacerbations among group B patients taking LABA/LAMA than among those taking LAMA monotherapy (odds ratio [OR], 0.258; 95% confidence interval [CI], 0.095–0.703). In contrast, in the subgroup of group D patients with COPD Assessment Test scores ≥25, compared with LAMA monotherapy, LABA/LAMA treatment was associated with lower risk of severe COPD exacerbations (OR, 0.115; 95% CI, 0.018-0.749). The combination of LABA and LAMA was found to be superior to LAMA monotherapy, especially for treating older adults with group B COPD, as well as for group D patients with severe symptoms.