• Journal of Korea Water Resources Association
• /
• v.57 no.9
• /
• pp.585-598
• /
• 2024

To mitigate natural disasters and efficiently manage water resources, it is essential to enhance hydrologic prediction while reducing model structural uncertainties. This study analyzed the impact of lumped and semi-distributed GR4J model structures on simulation performance and evaluated uncertainties with and without data assimilation techniques. The Ensemble Kalman Filter (EnKF) and Particle Filter (PF) methods were applied to the Namgang Dam basin. Simulation results showed that the Kling-Gupta efficiency (KGE) index was 0.749 for the lumped model and 0.831 for the semi-distributed model, indicating improved performance in semi-distributed modeling by 11.0%. Additionally, the impact of uncertainties in meteorological forcings (precipitation and potential evapotranspiration) on data assimilation performance was analyzed. Optimal uncertainty conditions varied by data assimilation method for the lumped model and by sub-basin for the semi-distributed model. Moreover, reducing the calibration period length during data assimilation led to decreased simulation performance. Overall, the semi-distributed model showed improved flood simulation performance when combined with data assimilation compared to the lumped model. Selecting appropriate hyper-parameters and calibration periods according to the model structure was crucial for achieving optimal performance.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.16-16
• /
• 2017

기후 시스템의 다양한 상호작용으로 인해 나타나는 대표적 현상인 강우는 수문학적 분석 과정의 필수적인 요소이며 장기 강우를 예측하는 것은 효율적인 수자원 관리에 중요한 기반이 되고 있다. 이러한 강우는 장기적으로 지구의 대기-해양 순환 패턴의 영향을 받으며, 특히 엘니뇨와 라니냐와 같은 기상 이변이 발생할 경우 대규모 순환에 변화가 일어나게 되어 강우에 영향을 미칠 수 있다. 따라서 본 연구에서는 지구의 순환 패턴 특성을 수치화한 전지구적 기상인자 중에서 우리나라 장기 강우를 예측하기 위한 기상인자를 선정하고 시계열 모형 구축을 통하여 예측력을 평가하였다. 이를 위해 강우에 내재된 다양한 대기-해양 순환 패턴으로부터 나타나는 주기적 요소를 추출하기 위해 앙상블 경험적 모드분해법을 사용하여 강우를 분해한 후, 각 분해된 강우자료와 전지구적 기상인자와의 상관성 분석을 통해 높은 상관성을 가진 기상인자를 선별하고 단계식 변수선택법으로부터 유의미한 기상인자를 최종적으로 선정하였다. 그 결과, 우리나라 기상청 60개 지점의 월별 강우자료 중 전반적으로 영향을 미치는 기상인자를 선정할 수 있었으며, 선정된 기상인 자로 구축된 시계열 모형을 통해 우리나라 장기 강우를 예측하였다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.168-168
• /
• 2018

최근 기후변화로 인하여 전 세계적으로 과거 강우사상에서 확인되지 않는 극치사상이 빈번하게 관측되고 있으며 이에 따른 피해도 증가하고 있다. 미래의 기상학적 변동성 및 기후변화 영향은 지구순환모형 (General Circulation Models, GCM)을 통해 구체화되며 가장 일반적인 기후변화 전망자료로서 활용된다. 그러나 산정된 기후변화 시나리오마다 서로 그 특성에 차이가 있으며 이러한 이유로 다양한 원인으로 인해 큰 변동성을 가지는 미래 극치강우를 하나의 시나리오로 분석하기에는 무리가 있다. 또한 다양한 시나리오를 통해 분석한 결과값이 상이하며 이러한 시나리오별 산정 결과의 차이는 사용자에게 혼란을 야기할 수 있어 이를 하나의 결과로 나타낼 필요성이 있으나 정량적인 대푯값을 얻기 위해 특정 시나리오를 선택하는 것은 신뢰성에 문제가 있다. 본 연구에서는 시나리오들을 정량적 지표에 의거하여 혼합된 하나의 시나리오로 표출하고자 하였다. CORDEX-RCMs 시나리오 중 HadGEM3-RA, RegCM, SNU_WRF 및 GRIMs를 입력 자료로 하여 다중모형앙상블(Multi-Model Ensemble, MME)을 통해 낙동강 유역의 극치강우에 대한 하나의 최적 기후변화 시나리오를 도출하고자 하였으며 계층적 베이지안 (Hierarchical Bayesian Model, HBM) 기법을 통하여 기후변화 시나리오에 내제된 불확실성에 대한 정량적인 해석을 수행하였다.

• Journal of Korea Water Resources Association
• /
• v.50 no.1
• /
• pp.17-28
• /
• 2017

A recent increase in extreme weather events and flash floods associated with the enhanced climate variability results in an increase in climate-related disasters. For these reasons, various studies based on a high resolution weather radar system have been carried out. The weather radar can provide estimates of precipitation in real-time over a wide area, while ground-based rain gauges only provides a point estimate in space. Weather radar is thus capable of identifying changes in rainfall structure as it moves through an ungauged basin. However, the advantage of the weather radar rainfall estimates has been limited by a variety of sources of uncertainty in the radar reflectivity process, including systematic and random errors. In this study, we developed an ensemble radar rainfall estimation scheme using the multivariate copula method. The results presented in this study confirmed that the proposed ensemble technique can effectively reproduce the rainfall statistics such as mean, variance and skewness (more importantly the extremes) as well as the spatio-temporal structure of rainfall fields.

• Journal of Cadastre & Land InformatiX
• /
• v.46 no.2
• /
• pp.43-55
• /
• 2016

This study estimated land prices using instance-based learning. A k-nearest neighbor method was utilized among various instance-based learning methods, and the 10 distance metrics including Euclidean distance were calculated in k-nearest neighbor estimation. One distance metric prediction which shows the best predictive performance would be normally chosen as final estimate out of 10 distance metric predictions. In contrast to this practice, an ensemble technique which combines multiple predictions to obtain better performance was applied in this study. We applied the gradient boosting algorithm, a sort of residual-fitting model to our data in ensemble combining. Sales price data of farm lands in Haenam-gun, Jeolla Province were used to demonstrate advantages of instance-based learning as well as an ensemble technique. The result showed that the ensemble prediction was more accurate than previous 10 distance metric predictions.

• Journal of Intelligence and Information Systems
• /
• v.19 no.2
• /
• pp.55-71
• /
• 2013

The prediction model is the main factor affecting the performance of a knowledge-based system for bankruptcy prediction. Earlier studies on prediction modeling have focused on the building of a single best model using statistical and artificial intelligence techniques. However, since the mid-1980s, integration of multiple techniques (hybrid techniques) and, by extension, combinations of the outputs of several models (ensemble techniques) have, according to the experimental results, generally outperformed individual models. An ensemble is a technique that constructs a set of multiple models, combines their outputs, and produces one final prediction. The way in which the outputs of ensemble members are combined is one of the important issues affecting prediction accuracy. A variety of combination schemes have been proposed in order to improve prediction performance in ensembles. Each combination scheme has advantages and limitations, and can be influenced by domain and circumstance. Accordingly, decisions on the most appropriate combination scheme in a given domain and contingency are very difficult. This paper proposes a confidence-based selection approach as part of an ensemble bankruptcy-prediction scheme that can measure unified confidence, even if ensemble members produce different types of continuous-valued outputs. The present experimental results show that when varying the number of models to combine, according to the creation type of ensemble members, the proposed combination method offers the best performance in the ensemble having the largest number of models, even when compared with the methods most often employed in bankruptcy prediction.

• The Korean Journal of Applied Statistics
• /
• v.27 no.3
• /
• pp.357-371
• /
• 2014

There are many studies related to imbalanced data in which the class distribution is highly skewed. To address the problem of imbalanced data, previous studies deal with resampling techniques which correct the skewness of the class distribution in each sampled subset by using under-sampling, over-sampling or hybrid-sampling such as SMOTE. Ensemble methods have also alleviated the problem of class imbalanced data. In this paper, we compare around a dozen algorithms that combine the ensemble methods and resampling techniques based on simulated data sets generated by the Backbone model, which can handle the imbalance rate. The results on various real imbalanced data sets are also presented to compare the effectiveness of algorithms. As a result, we highly recommend the resampling technique combining ensemble methods for imbalanced data in which the proportion of the minority class is less than 10%. We also find that each ensemble method has a well-matched sampling technique. The algorithms which combine bagging or random forest ensembles with random undersampling tend to perform well; however, the boosting ensemble appears to perform better with over-sampling. All ensemble methods combined with SMOTE outperform in most situations.

• Journal of Korea Society of Industrial Information Systems
• /
• v.25 no.2
• /
• pp.57-72
• /
• 2020

The price of iron ore has continued to fluctuate with high demand and supply from many countries and companies. In this business environment, forecasting the price of iron ore has become important. This study developed the machine learning model forecasting the price of iron ore a one month after the trading events. The forecasting model used distributed lag model and deep learning models such as MLP (Multi-layer perceptron), RNN (Recurrent neural network) and LSTM (Long short-term memory). According to the results of comparing individual models through metrics, LSTM showed the lowest predictive error. Also, as a result of comparing the models using the ensemble technique, the distributed lag and LSTM ensemble model showed the lowest prediction.

• Journal of Convergence for Information Technology
• /
• v.9 no.6
• /
• pp.83-90
• /
• 2019

People constantly strive for better choices. For this reason, recommender system has been developed since the early 1990s. In particular, collaborative filtering technique has shown excellent performance in the field of recommender systems, and research of recommender system using machine learning has been actively conducted. This study constructs recommender system using collaborative filtering and machine learning based on stacking model which is one of ensemble methods. The results of this study confirm that the recommender system with the stacking model is useful in aspects of recommender performance. In the future, the model proposed in this study is expected to help individuals or firms to make better choices.

• Journal of the Korean Data and Information Science Society
• /
• v.23 no.5
• /
• pp.949-959
• /
• 2012

This article is to find the right size of decision trees that performs better for boosting algorithm. First we defined the tree size D as the depth of a decision tree. Then we compared the performance of boosting algorithm with different tree sizes in the experiment. Although it is an usual practice to set the tree size in boosting algorithm to be small, we figured out that the choice of D has a significant influence on the performance of boosting algorithm. Furthermore, we found out that the tree size D need to be sufficiently large for some dataset. The experiment result shows that there exists an optimal D for each dataset and choosing the right size D is important in improving the performance of boosting. We also tried to find the model for estimating the right size D suitable for boosting algorithm, using variables that can explain the nature of a given dataset. The suggested model reveals that the optimal tree size D for a given dataset can be estimated by the error rate of stump tree, the number of classes, the depth of a single tree, and the gini impurity.