• Proceedings of the Korea Water Resources Association Conference
• /
• 2022.05a
• /
• pp.200-200
• /
• 2022

본 연구에서는 총 강우량과 강우강도을 고려한 이변수 분석으로 연최대 호우사상을 선별하고, 두 변수를 Copula 함수로 결합하여 최적의 모델조합을 찾는 확률호우사상 산정 방법론을 제시하였다. 국내 69개 관측소의 2020년까지의 관측 자료를 대상으로 1mm 이하의 강우는 제거한 뒤, IETD(Inter-Event Time Definition) 12시간을 기준으로 강우자료를 독립적인 호우사상으로 분리하였다. 호우사상의 여러 특성 중 양의 상관관계를 갖는 총 강우량과 강우강도를 변수로 선택해 이변수 지수분포에 대입하였고, 각 지점의 연최대 호우사상 시계열을 생성하였다. 2변수 지수분포의 매개변수는 전체 기간과 연도별로 나누어 추정해 본 결과 연도별 변동성이 큰 것을 확인해 연도별 추정 방식을 선택하였다. 연최대 강우사상 시계열의 총 강우량과 강우강도는 극한 강우에 적용하는 확률분포형 중 Lognarmal, Gamma, Gumbel, GEV(Generalized Extreme Value), GPD(Generalized Pareto Distribution) 5가지를 사용하여 각각 CDF(Cumulative distribution Function) 값을 추정하였다. 계산된 CDF 값은 3가지 Copula 모형으로 결합해 joint CDF 값을 산출하였다. 총 75개의 모델조합 중 최적 모델을 찾기 위해 CVM(Cramer-von-Mises) 적합도 검정을 시행하였다. CVM의 통계량 Sn 값이 가장 작은 모델조합을 해당 지점의 최적 모델조합으로 선정하였다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.1159-1163
• /
• 2010

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 요소이다. 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 이렇게 유도된 유출량의 경우 오차가 크기 때문에 그 신뢰성에 문제가 있을 것으로 판단된다. 따라서 본 논문에서는 선행우량 및 수위자료로부터 단시간 수위예측에 관해 연구하였다. 신경망은 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 연결강도를 최적화함으로서 모형의 구조를 스스로 조직화하기 때문에 모형의 구조에 적합한 최적 매개변수를 추정할 수 있다. 따라서 정확한 예측이 어려운 하천수위를 과거의 자료로 부터 학습된 신경망의 수학적 알고리즘을 통해 유출량의 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. 따라서 본 연구에서는 인공신경망의 가중치를 유전자 알고리즘에 의해 최적화시킨후 오류역전파알고리즘에 의해 신경망의 학습을 진행하는 모형으로 감천유역의 선산수위표지점의 수위를 1시간~6시간까지 예측하였다.

• Journal of Intelligence and Information Systems
• /
• v.10 no.1
• /
• pp.109-123
• /
• 2004

Corporate financial distress and bankruptcy prediction is one of the major application areas of artificial neural networks (ANNs) in finance and management. ANNs have showed high prediction performance in this area, but sometimes are confronted with inconsistent and unpredictable performance for noisy data. In addition, it may not be possible to train ANN or the training task cannot be effectively carried out without data reduction when the amount of data is so large because training the large data set needs much processing time and additional costs of collecting data. Instance selection is one of popular methods for dimensionality reduction and is directly related to data reduction. Although some researchers have addressed the need for instance selection in instance-based learning algorithms, there is little research on instance selection for ANN. This study proposes a genetic algorithm (GA) approach to instance selection in ANN for bankruptcy prediction. In this study, we use ANN supported by the GA to optimize the connection weights between layers and select relevant instances. It is expected that the globally evolved weights mitigate the well-known limitations of gradient descent algorithm of backpropagation algorithm. In addition, genetically selected instances will shorten the learning time and enhance prediction performance. This study will compare the proposed model with other major data mining techniques. Experimental results show that the GA approach is a promising method for instance selection in ANN.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.1605-1609
• /
• 2010

CGCM3.1 SRES B1 시나리오의 2D 변수들을 입력값으로 인공신경망 모형을 이용한 스케일 상세화기법으로 강부식(2009)은 소양강댐 유역의 월 누적강수 경향분석을 실시하였다. 원시 GCM 시나리오를 스케일 상세화 시키기 위한 기법의 하나로 인공신경망 모형을 사용할 수 있는데, 이 경우 GCM에서 모의되는 강수플럭스, 해면기압, 지표면 근처에서의 일 평균온도, 지표면 근처에서의 일평균온도, 지표면으로부터 발생하는 잠열플럭스 등과 같은 22개의 변수를 잠재적인 예측인자로 사용하여 신경망을 구성하게 된다. 입력변수세트의 구성은 인공신경망의 계산 효율을 좌우하는 중요한 요소라 할 수 있다. 본 연구에서는 변수의 물리적 특성을 고려하여 순차적인 변수선택을 통한 신경망 입력변수 세트를 구성하고 입력세트 간의 학습성과 비교를 통하여, 최적 입력변수 선정 및 신경망의 학습효과를 높일 수 있는 방법에 대해 연구하였다. 물리적 상관성이 높다고 판단되는 GCM_Prec, huss, ps를 입력변수로 하여 순차적인 케이스를 학습해본 결과 huss와 ps를 입력변수로 하는 케이스에 대해서 적은 오차와 높은 상관성을 보였다, 또한, 신경망의 학습 효과를 높이기 위해 홍수기와 비홍수기로 구분하여 학습한 결과 홍수기와 비홍수기로 구분하여 신경망을 구성하였을 경우가 향상된 모의값을 나타내었다. 기후변화모의자료는 CCCma(Canadian Center for Climate Modeling and Analysis)에서 제공되는 CGCM3.1/T63 20C3M 시나리오를 사용하였으며, 관측값으로는 AWS에서 제공된 일 누적강수를 사용하였다. 인공신경망의 학습기간은 1997년부터 2000년이며, 검증기간은 2001년부터 2004년으로 구성하였다.

• Journal of the Korea Society of Computer and Information
• /
• v.12 no.4
• /
• pp.43-52
• /
• 2007

As the number of public Web Services increases, there will be many services with the same functionality. These services. however, will vary in their QoS properties, such as price, response time and availability, and it is very important to choose a best service while satisfying given QoS constraints. This paper brings parallel branching and response time constraint of business processes into focus and investigates several service selection plans based on multidimensional multiple choice Knapsack model. Specifically. proposed in the paper are a plan with response time constraints for each execution flow, a plan with a single constraint over the whole service types and a plan with a constraint on a particular execution path of a composite Web Services. Experiments are conducted to observe the performance of each plan with varying the number of services, the number of branches and the values of response time constraint. Experimental results show that reducing the number of candidate services using Pareto Dominance is very effective and the plan with a constraint over the whole service types is efficient in time and solution quality for small to medium size problems.

• Journal of Korean Society of Transportation
• /
• v.29 no.1
• /
• pp.81-93
• /
• 2011

This research deals with the multi-modal continuous network design problem to resolve the transportation policy problems for constructing and operating transportation facilities with considering the mutual decision-making process between transportation operator and user in the multi-modal network. Particularly, in the consideration of changes in travel pattern between transport modes due to the changes in transportation policy, road network for passenger car and transit network for public transportation are considered together. In the development of network design model, more rational Stackelberg equilibrium(cooperative game) rather than more general Nash equilibrium(non-cooperative game) approach is used and sensitivity analysis considering transport mode is used. A multi-modal continuous network design model in this study is developed for the arbitrary continuous network design parameters(${\epsilon},\hat{\epsilon},p$) of transportation policy decisions. As examples of application and evaluation for these design parameters, the developed model is applied to calculate 1)the optimal capacity of road link in the road transport policy, 2)the optimal frequency of transit line in public transport policy and 3)the optimal modal split in transport modal share policy.

• Journal of the Korean Chemical Society
• /
• v.57 no.1
• /
• pp.147-158
• /
• 2013

This study is to search an optimal model on causal relationships of the motivations to learn and motivation strategy in chemistry education. The participants in this study are consisted of G and I high schools students (487) in Gwangju. They all answered to the questionnaire. Model I is hypothesized to be path model of the mediation between 'self-efficacy, achievement need, and entity theory of intelligence' and 'learning strategy and self-handicapping tendency of motivation strategy' by goal orientation to explore variables of study effecting the motivation strategy. And Model II is hypothesized path model of the mediation between goal orientation and 'learning strategy and self-handicapping tendency' by 'self-efficacy, achievement need, and entity theory' to explore variables of study effecting the motivation strategy. Based on these models, structural equation modeling techniques are used to evaluate for the path model among goal orientation(learning, performance approach, and performance approach goal orientation), self-efficacy, achievement need, entity theory of intelligence, self-handicapping tendency, and learning strategy in chemistry education. As the results, Model II is considered. Goodness-of-fit indexes of this model related modification models are identified and analyzed in phases. And this model is accomplished by correcting the model the fifth time to enhance goodness-of-fit indexes. In this optimal model II-5 (Fig. 3) on causal relationships of the motivations to learn and learning strategy (p

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.1 no.1
• /
• pp.52-57
• /
• 2002

A simulation tool for an optimal ATIS design and drivers' dynamic route choice behavior analysis is developed, which is applicable to urban networks. Due to the difficulty to make drivers feel the time pressure according to traffic conditions, current SP questionnaire survey type surveys have a limitation to capture correct driver reactions to real-time traffic Information provision. The simulator Is a web-based upgraded version, named WATiSim (Web-based ATIS Simulator), to quickly perform a wide population survey with a minimal cost using INTERNET Furthermore, the time pressure issue is lessened by its interface and simulation modules. After WATiSim mimicked a VMS based ATIS in a partial network of Seoul Metropolitan, reactions of drivers to various traffic conditions were surveyed through INTERNET and analyzed using a logit model. Drivers under the ATIS environment clearly understood the provided traffic information, and their reactions were closely related to traffic conditions, scheduled delay, trip purposes as well as toll charge if any.

• Journal of Korea Water Resources Association
• /
• v.39 no.12 s.173
• /
• pp.997-1012
• /
• 2006

We have presented a nonparametric stochastic approach for the SOI(Southern Oscillation Index) series that used nonlinear methodology called Nonlinear AutoRegressive(NAR) based on conditional kernel density function and CAFPE(Corrected Asymptotic Final Prediction Error) lag selection. The fitted linear AR model represents heteroscedasticity, and besides, a BDS(Brock - Dechert - Sheinkman) statistics is rejected. Hence, we applied NAR model to the SOI series. We can identify the lags 1, 2 and 4 are appropriate one, and estimated conditional mean function. There is no autocorrelation of residuals in the Portmanteau Test. However, the null hypothesis of normality and no heteroscedasticity is rejected in the Jarque-Bera Test and ARCH-LM Test, respectively. Moreover, the lag selection for conditional standard deviation function with CAFPE provides lags 3, 8 and 9. As the results of conditional standard deviation analysis, all I.I.D assumptions of the residuals are accepted. Particularly, the BDS statistics is accepted at the 95% and 99% significance level. Finally, we split the SOI set into a sample for estimating themodel and a sample for out-of-sample prediction, that is, we conduct the one-step ahead forecasts for the last 97 values (15%). The NAR model shows a MSEP of 0.5464 that is 7% lower than those of the linear model. Hence, the relevance of the NAR model may be proved in these results, and the nonparametric NAR model is encouraging rather than a linear one to reflect the nonlinearity of SOI series.

• Journal of Korea Water Resources Association
• /
• v.50 no.3
• /
• pp.191-200
• /
• 2017

A robust parameter set (ROPS) selection framework for an unsteady flow model was developed by combining Pareto optimums obtained by outcomes of model calibration using multi-site observations with the minimax regret approach (MRA). The multi-site calibration problem which is a multi-objective problem was solved by using an aggregation approach which aggregates the weighted criteria related to different sites into one measure, and then performs a large number of individual optimization runs with different weight combinations to obtain Pareto solutions. Roughness parameter structure which can describe the variation of Manning's n with discharges and sub-reaches was proposed and the related coefficients were optimized as model parameters. By applying the MRA which is a decision criterion, the Pareto solutions were ranked based on the obtained regrets related to each Pareto solution, and the top-rated one due to the lowest aggregated regrets of both calibration and validation was determined as the only ROPS. It was found that the determination of variable roughness and the corresponding standardized RMSEs at the two gauging stations varies considerably depending on the combinations of weights on the two sites. This method can provide the robust parameter set for the multi-site calibration problems in hydrologic and hydraulic models.