• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.25 no.5
• /
• pp.1-13
• /
• 2022

The damages caused by landslides are increasing worldwide due to climate change. In Korea, damages from landslides occur frequently, making it necessary to develop the effective response strategies. In particular, there is a lack of countermeasures against landslides in cultural heritage areas. The purpose of this study was to spatially analyze the relationship between Buyeo-gun's cultural heritage and landslide susceptible areas in Buyeo-gun, Chungcheongnam-do, which has a long history. Nine spatial distribution models were used to evaluate the landslide susceptibility, and the ensemble method was applied to reduce the uncertainty of individual model. There were 17 cultural heritages belonging to the landslide susceptible area. As a result of calculating the area ratio of the landslide susceptible area for cultural heritages, the cultural heritages with 100% of the area included in the landslide susceptible area were "Standing statue of Maae in Hongsan Sangcheon-ri" and "Statue of King Seonjo." More than 35% of "Jeungsanseong", "Garimseong", and "Standing stone statue of Maitreya Bodhisattva in Daejosa Temple" belonged to landslide susceptible areas. In order to effectively prevent landslide damage, the application of landslide prevention measures should be prioritized according to the proportion belonging to the landslide susceptible area. Since it is very difficult to restore cultural properties once destroyed, preventive measures are required before landslide damage occurs. The approach and results of this study provide basic data and guidelines for disaster response plans to prevent landslides in Buyeo-gun.

• Korea Trade Review
• /
• v.48 no.2
• /
• pp.27-43
• /
• 2023

The maritime industry is playing an increasingly vital part in global economic expansion. Specifically, the Baltic Dry Index is highly correlated with global commodity prices. Hence, the importance of BDI prediction research increases. But, since the global situation has become more volatile, it has become methodologically more difficult to predict the BDI accurately. This paper proposes an integrated machine-learning strategy for accurately forecasting BDI trends. This study combines the benefits of a convolutional neural network (CNN) and long short-term memory neural network (LSTM) for research on prediction. We collected daily BDI data for over 27 years for model fitting. The research findings indicate that CNN successfully extracts BDI data features. On this basis, LSTM predicts BDI accurately. Model R2 attains 94.7 percent. Our research offers a novel, machine-learning-integrated approach to the field of shipping economic indicators research. In addition, this study provides a foundation for risk management decision-making in the fields of shipping institutions and financial investment.

• Journal of the Korean Geosynthetics Society
• /
• v.23 no.1
• /
• pp.17-25
• /
• 2024

The Voids in the Mineral Aggregate (VMA) within asphalt mixtures play a crucial role in defining the mixture's structural integrity, durability, and resistance to environmental factors. Accurate prediction and optimization of VMA are essential for enhancing the performance and longevity of asphalt pavements, particularly in varying climatic and environmental conditions. This study introduces a novel machine learning framework leveraging ensemble machine learning model for predicting VMA in asphalt mixtures. By analyzing a comprehensive set of variables, including aggregate size distribution, binder content, and compaction levels, our framework offers a more precise prediction of VMA than traditional single-model approaches. The use of advanced machine learning techniques not only surpasses the accuracy of conventional empirical methods but also significantly reduces the reliance on extensive laboratory testing. Our findings highlight the effectiveness of a data-driven approach in the field of asphalt mixture design, showcasing a path toward more efficient and sustainable pavement engineering practices. This research contributes to the advancement of predictive modeling in construction materials, offering valuable insights for the design and optimization of asphalt mixtures with optimal void characteristics.

• Interaction and multiscale mechanics
• /
• v.1 no.2
• /
• pp.303-315
• /
• 2008

In this paper, thermodynamical properties of crystalline silicon under strain are calculated using classical molecular dynamics (MD) simulations based on the Tersoff interatomic potential. The Helmholtz free energy of the silicon crystal under strain is calculated by using the ensemble method developed by Frenkel and Ladd (1984). To account for quantum corrections under strain in the classical MD simulations, we propose an approach where the quantum corrections to the internal energy and the Helmholtz free energy are obtained by using the corresponding energy deviation between the classical and quantum harmonic oscillators. We calculate the variation of thermodynamic properties with temperature and strain and compare them with results obtained by using the quasi-harmonic model in the reciprocal space.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.5
• /
• pp.138-151
• /
• 1996

This paper proposes a genetic fuzzy controller ensemble (FCE) for improving the control performance of of fuzzy controller in the non-linear and complex problems. The design procedure of each fuzzy controller in the FCF consists of the following two stages, each of which is performed by different genetic algorithms. The first stage generates a fuzzy rule base that covers the training examples as many as possible. The second stage builds fine-tuned membership funcitons that make the control error as small as possible. These two stages are repeated independently upon the different partition patterns of input-output variables. The control performance of the proposed method is compared with that of wang and mendel's approach[1] in terms of either the percentage of successful controls reaching to the goal or the average traveling distance.

• Proceedings of the KIEE Conference
• /
• 2005.07a
• /
• pp.834-836
• /
• 2005

This paper describes algorithms for enhancing the reliability of measurements to verify the performance of energy efficiency programs with an simple method. Fundamentally, measurements contain erroneous data because of the various causes. and so proper procedures or schemes are prepared before the performance is evaluated. In this paper, we propose an approach for detecting and correcting an adulterate data, such as missing and bad data. Erroneous data are detected or corrected by pre-described measuring conditions, ensemble average, and standard deviation of measurements at measuring time. The proposed algorithms are tested by field test measurements. From case studies we drew the promising results.

• The Journal of Information Systems
• /
• v.20 no.3
• /
• pp.237-255
• /
• 2011

지난 수십 년 간 연구자들은 효과적인 진료지원시스템을 개발하기 위해 다양한 도구와 방법론들을 제안하였고 지금도 새로운 방법론과 도구들을 계속적으로 개발하고 있다. 그 중에서 흉통으로 응급실에 내원한 노인환자에 대한 정확한 진단은 중요한 이슈 중의 하나였다. 따라서 많은 연구자들이 의사의 진단 능력을 향상시키기 위한 지능적인 의료의사결정과 시스템 개발에 투신하고 있지만 전통적인 의료시스템에 따른 대부분의 진료의사결정이 단일 분류기(classifier)에 기반하고 있어 만족스런 성능을 보여주지 못하고 있는 것이 현실이다. 따라서 이 논문은 앙상블 전략을 활용하여 의사들이 노인환자들의 흉통을 더 정확하고 빠르게 진단하는데 있어 도움을 줄 수 있게 하였다. 의사결정나무, 인공신경망, SVM 모델을 결합한 앙상블 기법을 실제 응급실에서 수집한 응급실 자료에 적용하였고, 그 결과 단일 분류기를 사용하는 것에 비해 월등히 향상된 진단 성과를 보이는 것을 관찰 할 수 있었다.

• International Journal of Contents
• /
• v.7 no.3
• /
• pp.1-5
• /
• 2011

It has been an interesting challenge to find a good classifier for imbalanced data, since it is pervasive but a difficult problem to solve. However, classifiers developed with the assumption of well-balanced class distributions show poor classification performance for the imbalanced data. Among many approaches to the imbalanced data problems, the algorithmic level approach is attractive because it can be applied to the other approaches such as data level or ensemble approaches. Especially, the error back-propagation algorithm using the target node method, which can change the amount of weight-updating with regards to the target node of each class, attains good performances in the imbalanced data problems. In this paper, we analyze the relationship between two optimal outputs of neural network classifier trained with the target node method. Also, the optimal relationship is compared with those of the other error function methods such as mean-squared error and the n-th order extension of cross-entropy error. The analyses are verified through simulations on a thyroid data set.

• Wind and Structures
• /
• v.12 no.2
• /
• pp.133-149
• /
• 2009

This paper outlines the results of a physical simulation (at a 1:700 - 1:1000 geometric scale) of a thunderstorm downburst. Three different methods are examined in order to generate the time dependent nature of a downburst: directly controlling the fans and via two different types of opening apertures. Similarities are shown to exist between each method, although the results obtained from one approach are favoured since they appear to be independent of the downdraft velocity. Significant run-to-run variations between each experiment are discovered and in general it is found beneficial to interpret the results in terms of 10 run ensemble averages. An attempt to simulate a translating downburst is also undertaken and the results are shown to compare favourably with full-scale data.

• International journal of advanced smart convergence
• /
• v.9 no.1
• /
• pp.193-201
• /
• 2020

The main objective of this study is to investigate the impact of additional modalities on the performance of emotion recognition using speech, facial expression and physiological measurements. In order to compare different approaches, we designed a feature-based recognition system as a benchmark which carries out linear supervised classification followed by the leave-one-out cross-validation. For the classification of four emotions, it turned out that bimodal fusion in our experiment improves recognition accuracy of unimodal approach, while the performance of trimodal fusion varies strongly depending on the individual. Furthermore, we experienced extremely high disparity between single class recognition rates, while we could not observe a best performing single modality in our experiment. Based on these observations, we developed a novel fusion method, called parametric decision fusion (PDF), which lies in building emotion-specific classifiers and exploits advantage of a parametrized decision process. By using the PDF scheme we achieved 16% improvement in accuracy of subject-dependent recognition and 10% for subject-independent recognition compared to the best unimodal results.