• 한국전산유체공학회:학술대회논문집
• /
• 2008.10a
• /
• pp.42-49
• /
• 2008

A wavelet method is presented in order to improve the computational efficiency of two dimensional unsteady flow problems while maintaining the order of accuracy of conventional CFD schemes. First, by using the interpolating wavelet transformation including decomposition and thresholding, an adaptive dataset to a solution is constructed. Then, inviscid and viscous fluxes are calculated only at the points within an adaptive dataset, which enhances the computational efficiency. Second, thresholding step is modified to maintain the spatial and temporal accuracy of conventional CFD schemes automatically by selecting the threshold value between user-defined value and the magnitude of spatial or temporal truncation error. The wavelet method suggested in this study is successfully applied to various unsteady flow problems and it is shown that the computational efficiency is enhanced with maintaining the computational accuracy of CFD schemes.

• Food Science of Animal Resources
• /
• v.37 no.1
• /
• pp.139-146
• /
• 2017

Response surface methodology (RSM) is a useful set of statistical techniques for modeling and optimizing responses in research studies of food science. In the analysis of response surface data, a second-order polynomial regression model is usually used. However, sometimes we encounter situations where the fit of the second-order model is poor. If the model fitted to the data has a poor fit including a lack of fit, the modeling and optimization results might not be accurate. In such a case, using a fullest balanced model, which has no lack of fit, can fix such problem, enhancing the accuracy of the response surface modeling and optimization. This article presents how to develop and use such a model for the better modeling and optimizing of the response through an illustrative re-analysis of a dataset in Park et al. (2014) published in the Korean Journal for Food Science of Animal Resources.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.7
• /
• pp.75-82
• /
• 2017

Correct prediction of emotion is essential for developing advanced health devices. For this purpose, neural network has been successfully used. However, interpretation of how a certain emotion is predicted through the emotion prediction neural network is very tough. When interpreting mechanism about how emotion is predicted by using the emotion prediction neural network can be developed, such mechanism can be effectively embedded into highly advanced health-care devices. In this sense, this study proposes a novel approach to interpreting how the emotion prediction neural network yields emotion. Our proposed mechanism is based on HRV (heart rate variability) measurements, which is based on calculating physiological data out of ECG (electrocardiogram) measurements. Experiment dataset with 23 qualified participants were used to obtain the seven HRV measurement such as Mean RR, SDNN, RMSSD, VLF, LF, HF, LF/HF. Then emotion prediction neural network was modelled by using the HRV dataset. By applying the proposed mechanism, a set of explicit mathematical functions could be derived, which are clearly and explicitly interpretable. The proposed mechanism was compared with conventional neural network to show validity.

• Journal of the Korean Solar Energy Society
• /
• v.38 no.1
• /
• pp.1-11
• /
• 2018

In this work, we propose weather prediction models to estimate hourly outdoor temperatures and solar irradiance in the next day using forecasting information. Hourly weather data predicted by the proposed models are useful for setting system operating strategies for the next day. The outside temperature prediction model considers 3-hourly temperatures forecasted by Korea Meteorological Administration. Hourly data are obtained by a simple interpolation scheme. The solar irradiance prediction is achieved by constructing a dataset with the observed cloudiness and correspondent solar irradiance during the last two weeks and then by matching the forecasted cloud factor for the next day with the solar irradiance values in the dataset. To verify the usefulness of the weather prediction models in predicting a short-term building load, the predicted data are inputted to a TRNSYS building model, and results are compared with a reference case. Results show that the test case can meet the acceptance error level defined by the ASHRAE guideline showing 8.8% in CVRMSE in spite of some inaccurate predictions for hourly weather data.

• ETRI Journal
• /
• v.33 no.5
• /
• pp.741-751
• /
• 2011

This paper focuses on online speaker tracking for telephone conversations and broadcast news. Since the online applicability imposes some limitations on the tracking strategy, such as data insufficiency, a reliable approach should be applied to compensate for this shortage. In this framework, a set of reference speaker models are used as side information to facilitate online tracking. To improve the indexing accuracy, adaptation approaches in eigenvoice decomposition space are proposed in this paper. We believe that the eigenvoice adaptation techniques would help to embed the speaker space in the models and hence enrich the generality of the selected speaker models. Also, an index structure of the reference models is proposed to speed up the search in the model space. The proposed framework is evaluated on 2002 Rich Transcription Broadcast News and Conversational Telephone Speech corpus as well as a synthetic dataset. The indexing errors of the proposed framework on telephone conversations, broadcast news, and synthetic dataset are 8.77%, 9.36%, and 12.4%, respectively. Using the index tree structure approach, the run time of the proposed framework is improved by 22%.

• Industrial Engineering and Management Systems
• /
• v.9 no.2
• /
• pp.141-156
• /
• 2010

Despite all attempts to prevent fraud, it continues to be a major threat to industry and government. Traditionally, organizations have focused on fraud prevention rather than detection, to combat fraud. In this paper we present a role mining inspired approach to represent user behaviour in Enterprise Resource Planning (ERP) systems, primarily aimed at detecting opportunities to commit fraud or potentially suspicious activities. We have adapted an approach which uses set theory to create transaction profiles based on analysis of user activity records. Based on these transaction profiles, we propose a set of (1) anomaly types to detect potentially suspicious user behaviour, and (2) scenarios to identify inadequate segregation of duties in an ERP environment. In addition, we present two algorithms to construct a directed acyclic graph to represent relationships between transaction profiles. Experiments were conducted using a real dataset obtained from a teaching environment and a demonstration dataset, both using SAP R/3, presently the predominant ERP system. The results of this empirical research demonstrate the effectiveness of the proposed approach.

• Journal of the Korean Institute of Educational Facilities
• /
• v.26 no.5
• /
• pp.3-10
• /
• 2019

This study has a purpose to profile local sectors into meaningful groups by using facilities rates of Social Overhead Capital(SOC) for daily life. Comparing SOC for daily life among the meaningful groups, the profiling and comparison results bring the comprehensive understanding about the educational facilities in local sectors. For the research purpose, this study utilized Latent Profile Analysis(LPA) by using variables such as population, road information, SOC for daily life, usage of land, possession of land, and appraised value of land from the 2018 Geographic Information System(GIS) dataset of Gwangjin-gu, where is one of the administrative district of Seoul City. Results showed that there are four latent groups of sectors among 904 local sectors(100 squared-meters sector per each) in Gwangjin-gu. By comparing the four latent groups by using LPA, the results diagnose each sector's status and help to improve the policy about educational facilities. Specifically, by using dataset for SOC of daily life, there are four groups of local sectors and each group has different features. Based on the different features of local sector groups, there can be improved management of educational facilities matching with each group's features.

• Journal of Multimedia Information System
• /
• v.6 no.2
• /
• pp.81-86
• /
• 2019

Deep learning is emerging as one of the best tool in processing data related to medical imaging. In our research work, we have proposed a deep learning based framework CNN (Convolutional Neural Network) for the classification of dysplastic tissue images. The CNN has classified the given images into 4 different classes namely normal tissue, mild dysplastic tissue, moderate dysplastic tissue and severe dysplastic tissue. The dataset under taken for the study consists of 672 tissue images of epithelial squamous layer of oral cavity captured out of the biopsy samples of 52 patients. After applying the data pre-processing and augmentation on the given dataset, 2688 images were created. Further, these 2688 images were classified into 4 categories with the help of expert Oral Pathologist. The classified data was supplied to the convolutional neural network for training and testing of the proposed framework. It has been observed that training data shows 91.65% accuracy whereas the testing data achieves 89.3% accuracy. The results produced by our proposed framework are also tested and validated by comparing the manual results produced by the medical experts working in this area.

• Ocean Systems Engineering
• /
• v.10 no.4
• /
• pp.415-434
• /
• 2020

Detecting objects is important for the safe operation of ships, and enables collision avoidance, risk detection, and autonomous sailing. This study proposes a ship detection method from images and videos taken at sea using one of the state-of-the-art deep neural network-based object detection algorithms. A deep learning model is trained using a public maritime dataset, and results show it can detect all types of floating objects and classify them into ten specific classes that include a ship, speedboat, and buoy. The proposed deep learning model is compared to a universal trained model that detects and classifies objects into general classes, such as a person, dog, car, and boat, and results show that the proposed model outperforms the other in the detection of maritime objects. Different deep neural network structures are then compared to obtain the best detection performance. The proposed model also shows a real-time detection speed of approximately 30 frames per second. Hence, it is expected that the proposed model can be used to detect maritime objects and reduce risks while at sea.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.12
• /
• pp.4664-4681
• /
• 2020

Modulation recognition (MR) plays a key role in cognitive radar, cognitive radio, and some other civilian and military fields. While existing methods can identify the signal modulation type by extracting the signal characteristics, the quality of feature extraction has a serious impact on the recognition results. In this paper, an end-to-end MR method based on long short-term memory (LSTM) and the gated recurrent unit (GRU) is put forward, which can directly predict the modulation type from a sampled signal. Additionally, the sliding window method is applied to fast-changing mixed-modulation signals for which the signal modulation type changes over time. The recognition accuracy on training datasets in different SNR ranges and the proportion of each modulation method in misclassified samples are analyzed, and it is found to be reasonable to select the evenly-distributed and full range of SNR data as the training data. With the improvement of the SNR, the recognition accuracy increases rapidly. When the length of the training dataset increases, the neural network recognition effect is better. The loss function value of the neural network decreases with the increase of the training dataset length, and then tends to be stable. Moreover, when the fast-changing period is less than 20ms, the error rate is as high as 50%. As the fast-changing period is increased to 30ms, the error rates of the GRU and LSTM neural networks are less than 5%.