• Proceedings of the Korea Water Resources Association Conference
• /
• 2015.05a
• /
• pp.606-609
• /
• 2015

본 연구에서는 support vector regression (SVR) 및 매개변수 최적화 알고리즘을 이용한 하천수위 예측모델을 구축하고 이를 실제 유역에 적용하여 모델 효율성을 평가하였다. 여기서, SVR은 하천수위를 예측하기 위한 예측모델로서 채택되었으며, 커널함수 (Kernel function)로서는 radial basis function (RBF)을 선택하였다. 최적화 알고리즘은 SVR의 최적 매개변수 (C?, cost parameter or regularization parameter; ${\gamma}$, RBF parameter; ${\epsilon}$, insensitive loss function parameter)를 탐색하기 위하여 적용되었다. 매개변수 최적화 알고리즘으로는 grid search (GS), genetic algorithm (GA), particle swarm optimization (PSO), artificial bee colony (ABC) 알고리즘을 채택하였으며, 비교분석을 통해 최적화 알고리즘의 적용성을 평가하였다. 또한 SVR과 최적화 알고리즘을 결합한 모델 (SVR-GS, SVR-GA, SVR-PSO, SVR-ABC)은 기존에 수자원 분야에서 널리 적용되어온 신경망(Artificial neural network, ANN) 및 뉴로퍼지 (Adaptive neuro-fuzzy inference system, ANFIS) 모델과 비교하였다. 그 결과, 모델 효율성 측면에서 SVR-GS, SVR-GA, SVR-PSO 및 SVR-ABC는 ANN보다 우수한 결과를 나타내었으며, ANFIS와는 비슷한 결과를 나타내었다. 또한 SVR-GA, SVR-PSO 및 SVR-ABC는 SVR-GS보다 상대적으로 우수한 결과를 나타내었으며, 모델 효율성 측면에서 SVR-PSO 및 SVR-ABC는 가장 우수한 모델 성능을 나타내었다. 따라서 본 연구에서 적용한 매개변수 최적화 알고리즘은 SVR의 매개변수를 최적화하는데 효과적임을 확인할 수 있었다. SVR과 최적화 알고리즘을 이용한 하천수위 예측모델은 기존의 ANN 및 ANFIS 모델과 더불어 하천수위 예측을 위한 효과적인 도구로 사용될 수 있을 것으로 판단된다.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2016.05a
• /
• pp.298-298
• /
• 2016

최근 기후변화의 영향으로 시간에 따라 자료 및 통계적 특성이 변하는 비정상성이 다양한 수문자료에서 관측됨에 따라 비정상성 빈도해석에 대한 연구가 활발히 진행되고 있다. 비정상성 빈도해석에 사용되는 비정상성 확률 모형은 기존의 매개변수를 시간에 따라 변하는 공변량이 포함된 함수의 형태로 나타내기 때문에, 정상성 확률 모형에 비해 매개변수의 개수가 많으며 복잡한 형태를 가지게 된다. 따라서 본 연구에서는 비정상성 고려 시 모형이 복잡해짐에 따라 매개변수 및 확률 수문량의 불확실성이 어떻게 변하는지 알아보고자 하였다. 베이지안 방법은 매개변수 추정 및 확률 수문량의 산정 뿐 아니라 이에 대한 불확실성을 정량화할 수 있는 방법 중 하나이다. 따라서 베이지안 방법에서 매개변수 추정에 주로 쓰이는 Monte Carlo Markov Chain (MCMC) 방법 중 하나인 Metropolis-Hastings 알고리즘을 이용하여 정상성 및 비정상성 GEV모형에 대한 매개변수 및 확률수문량의 사후분포를 산정하였다. 산정된 사후분포의 사후구간을 통해 각 모형의 불확실성을 정량화하였으며, 계산된 불확실성의 비교를 통해 모형의 복잡성이 불확실성에 미치는 영향을 평가하였다.

• Journal of Intelligence and Information Systems
• /
• v.28 no.3
• /
• pp.45-55
• /
• 2022

Recently, large-scale language models (LPLM) have been shown state-of-the-art performances in various tasks of natural language processing including intent classification. However, fine-tuning LPLM requires much computational cost for training and inference which is not appropriate for dialog system. In this paper, we propose compressed intent classification model for multi-agent in low-resource like CPU. Our method consists of two stages. First, we trained sentence encoder from LPLM then compressed it through knowledge distillation. Second, we trained agent-specific adapter for intent classification. The results of three intent classification datasets show that our method achieved 98% of the accuracy of LPLM with only 21% size of it.

• Steel and Composite Structures
• /
• v.43 no.5
• /
• pp.625-637
• /
• 2022

Axial compression capacity (P_u) is a significant yet complex parameter of concrete-filled steel tube (CFST) columns. This study offers a novel ensemble tool, adaptive neuro-fuzzy inference system (ANFIS) supervised by equilibrium optimization (EO), for accurately predicting this parameter. Moreover, grey wolf optimization (GWO) and Harris hawk optimizer (HHO) are considered as comparative supervisors. The used data is taken from earlier literature provided by finite element analysis. ANFIS is trained by several population sizes of the EO, GWO, and HHO to detect the best configurations. At a glance, the results showed the competency of such ensembles for learning and reproducing the P_u behavior. In details, respective mean absolute errors along with correlation values of 4.1809% and 0.99564, 10.5947% and 0.98006, and 4.8947% and 0.99462 obtained for the EO-ANFIS, GWO-ANFIS, and HHO-ANFIS, respectively, indicated that the proposed EO-ANFIS can analyze and predict the behavior of CFST columns with the highest accuracy. Considering both time and accuracy, the EO provides the most efficient optimization of ANFIS and can be a nice substitute for experimental approaches.

• Journal of Korea Multimedia Society
• /
• v.25 no.2
• /
• pp.440-449
• /
• 2022

In recent years, the air pollution and Air Quality Index (AQI) has been a pivotal point for researchers due to its effect on human health. Various research has been done in predicting the AQI but most of these studies, either lack dense temporal data or cover one or two air pollutant elements. In this paper, a hybrid Convolutional Neural approach integrated with recurrent neural network architecture (CNN-LSTM), is presented to find air pollution inference using a multivariate air pollutant elements dataset. The aim of this research is to design a robust and real-time air pollutant forecasting system by exploiting a neural network. The proposed approach is implemented on a 24-month dataset from Seoul, Republic of Korea. The predicted results are cross-validated with the real dataset and compared with the state-of-the-art techniques to evaluate its robustness and performance. The proposed model outperforms SVM, SVM-Polynomial, ANN, and RF models with 60.17%, 68.99%, 14.6%, and 6.29%, respectively. The model performs SVM and SVM-Polynomial in predicting O3 by 78.04% and 83.79%, respectively. Overall performance of the model is measured in terms of Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and the Root Mean Square Error (RMSE).

• The Bulletin of The Korean Astronomical Society
• /
• v.44 no.2
• /
• pp.63.3-63.3
• /
• 2019

Nearby young moving groups (NYMGs hereafter) are gravitationally unbound loose young stellar associations located within 100 pc of the Sun. Since NYMGs are crucial laboratories for studying low-mass stars and planets, intensive searches for NYMG members have been performed. For identification of NYMG members, various strategies and methods have been applied. As a result, the reliability of the members in terms of membership is not uniform, which means that a careful membership re-assessment is required. In this study, I developed a NYMG membership probability calculation tool based on Bayesian inference (Bayesian Assessment of Moving Groups: BAMG). For the development of the BAMG tool, I constructed ellipsoidal models for nine NYMGs via iterative and self-consistent processes. Using BAMG, memberships of claimed members in the literature (N~2000) were evaluated, and 35 per cent of members were confirmed as bona fide members of NYMGs. Based on the deficiency of low-mass members appeared in mass function using these bona fide members, low mass members from Gaia DR2 are identified. About 2000 new M dwarf and brown dwarf candidate members were identified. Memberships of ~70 members with RV from Gaia were confirmed, and the additional ~20 members were confirmed via spectroscopic observation. Not relying on previous knowledge about the existence of nine NYMGs, unsupervised machine learning analyses were applied to NYMG members. K-means and Agglomerative Clustering algorithms result in similar trends of grouping. As a result, six previously known groups (TWA, beta-Pic, Carina, Argus, AB Doradus, and Volans-Carina) were rediscovered. Three the other known groups are recognized as well; however, they are combined into two new separate groups (ThOr+Columba and TucHor+Columba).

• The Korean Journal of Applied Statistics
• /
• v.33 no.1
• /
• pp.47-59
• /
• 2020

In this paper, we present the longitudinal Markov binary regression model with t-link function when its transition order is known or unknown. It is assumed that logit or probit models are considered in binary regression models. Here, t-link function can be used for more flexibility instead of the probit model since the t distribution approaches to normal distribution as the degree of freedom goes to infinity. A Markov regression model is considered because of the longitudinal data of each individual data set. We propose Bayesian method to determine the transition order of Markov regression model. In particular, we use the deviance information criterion (DIC) (Spiegelhalter et al., 2002) of possible models in order to determine the transition order of the Markov binary regression model if the transition order is known; however, we compute and compare their posterior probabilities if unknown. In order to overcome the complicated Bayesian computation, our proposed model is reconstructed by the ideas of Albert and Chib (1993), Kuo and Mallick (1998), and Erkanli et al. (2001). Our proposed method is applied to the simulated data and real data examined by Sommer et al. (1984). Markov chain Monte Carlo methods to determine the optimal model are used assuming that the transition order of the Markov regression model are known or unknown. Gelman and Rubin's method (1992) is also employed to check the convergence of the Metropolis Hastings algorithm.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.30 no.2
• /
• pp.147-156
• /
• 2024

The sustainable management and enhancement of marine resources are becoming increasingly important issues worldwide. This study was conducted in response to these challenges, focusing on the development and performance comparison of fish detection and classification models as part of a deep learning-based technique for assessing the effectiveness of marine resource enhancement projects initiated by the Korea Fisheries Resources Agency. The aim was to select the optimal model by training various sizes of YOLOv8-Seg models on a fish image dataset and comparing each performance metric. The dataset used for model construction consisted of 36,749 images and label files of 12 different species of fish, with data diversity enhanced through the application of augmentation techniques during training. When training and validating five different YOLOv8-Seg models under identical conditions, the medium-sized YOLOv8m-Seg model showed high learning efficiency and excellent detection and classification performance, with the shortest training time of 13 h and 12 min, an of 0.933, and an inference speed of 9.6 ms. Considering the balance between each performance metric, this was deemed the most efficient model for meeting real-time processing requirements. The use of such real-time fish detection and classification models could enable effective surveys of marine resource enhancement projects, suggesting the need for ongoing performance improvements and further research.

• Journal of the Korean Chemical Society
• /
• v.54 no.3
• /
• pp.338-349
• /
• 2010

This study compared the instruction effectiveness between experiments of precipitation and a ball & stick model related to the 'Law of Definite Proportions' of 9th grade science in this study. The subjects were 250 students in the 9th grade. They were divided into two groups, an experimental group and a model group. The results showed that the ratio of thought in which the elements were divided in the solutions and the ratio of thought in which a new compound was created when the two solutions were mixed were higher in the precipitation experiment group than in the model group. The two groups were not different in terms of the ratio of thought related to the reason for the creation of the precipitate. The ratio of thought pertaining to incorrect answers was high, implying that the two strategies were not effective in correcting students' thoughts. However, the ratio of finding patterns from the measuring data in the model group was higher than in the experimental group. However, the ratios of 'definite proportions' inference in the bonding of the reactants were similar in the two groups. From these data, we concluded that the inference of the 'Law of definite proportion' from experiments or models was not suitable for middle school students.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.31 no.5B
• /
• pp.405-414
• /
• 2011

The adaptive network-based fuzzy inference system (ANFIS) which had a success for time series prediction and system control was applied for modeling the hourly runoff in the Gapcheon watershed. The ANFIS used the antecedent rainfall and runoff as the input. The ANFIS was trained by varying the various simulation factors such as mean areal rainfall estimation, the number of input variables, the type of membership function and the number of membership function. The root mean square error (RMSE), mean peak runoff error (PE), and mean peak time error (TE) were used for validating the ANFIS simulation. The ANFIS predicted runoff was in good agreement with the measured runoff and the applicability of ANFIS for modelling the hourly runoff appeared to be good. The forecasting ability of ANFIS up to the maximum 8 lead hour was investigated by applying the different input structure to ANFIS model. The accuracy of ANFIS for predicting the hourly runoff was reduced as the forecasting lead hours increased. The long-term predictability of ANFIS for forecasting the hourly runoff at longer lead hours appeared to be limited. The ANFIS might be useful for modeling the hourly runoff and has an advantage over the physically based models because the model construction of ANFIS based on only input and output data is relatively simple.