• Journal of information and communication convergence engineering
• /
• v.21 no.4
• /
• pp.294-299
• /
• 2023

Micro ribonucleic acids (miRNAs) can regulate the protein expression levels of genes in the human body and have recently been reported to be closely related to the cause of disease. Determining the genes related to miRNAs will aid in understanding the mechanisms underlying complex miRNAs. However, the identification of miRNA-related genes through wet experiments (in vivo, traditional methods are time- and cost-consuming). To overcome these problems, recent studies have investigated the prediction of miRNA relevance using deep learning models. This study presents a method for predicting the relationships between miRNAs and genes. First, we reconstruct a negative dataset using the proposed method. We then extracted the feature using an autoencoder, after which the feature vector was concatenated with the original data. Thereafter, the concatenated data were used to train a long short-term memory model. Our model exhibited an area under the curve of 0.9609, outperforming previously reported models trained using the same dataset.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.40 no.8
• /
• pp.711-719
• /
• 2016

In this study, we derived an theoretical equation, using a simplified spring-mass model for the rolling stock, to obtain the overriding behavior of a leading vehicle, which is considered as the main factor in train accidents. To verify the derived equation, we created a simple 2D model based on the theoretical model, and a simple 3D model considering the characteristics of the power bogie. We then compared the theoretical results with the simulation results obtained using LS-DYNA. The maximum relative derivations in the vertical displacements at the first end-buffer, which is the most important point in overriding, were 3.5 [%] and 1.7 [%] between the two results. Further, we evaluated collision-induced overriding displacements using the theoretical equation for a rubber draft gear, a hydraulic buffer under various collision conditions. We have suggested a theoretical approach for the realization of overriding collision accidents or the energy absorption design of the front end of trains.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.19-31
• /
• 2022

In Machine learning, we usually divide the entire data into training data and test data, train the model using training data, and use test data to determine the accuracy and generalization performance of the model. In the case of models with low generalization performance, the prediction accuracy of newly data is significantly reduced, and the model is said to be overfit. This study is about a method of generating training data based on central limit theorem and combining it with existed training data to increase normality and using this data to train models and increase generalization performance. To this, data were generated using sample mean and standard deviation for each feature of the data by utilizing the characteristic of central limit theorem, and new training data was constructed by combining them with existed training data. To determine the degree of increase in normality, the Kolmogorov-Smirnov normality test was conducted, and it was confirmed that the new training data showed increased normality compared to the existed data. Generalization performance was measured through differences in prediction accuracy for training data and test data. As a result of measuring the degree of increase in generalization performance by applying this to K-Nearest Neighbors (KNN), Logistic Regression, and Linear Discriminant Analysis (LDA), it was confirmed that generalization performance was improved for KNN, a non-parametric technique, and LDA, which assumes normality between model building.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.17 no.2
• /
• pp.57-66
• /
• 2009

In this paper, the steering characteristics of tracked vehicles are studied for the improvement of steering performance. The important design factor of military vehicles is high mobility. It is influenced by weight of a vehicle, engine capacity, power-train, and steering system. The military vehicle, which is equipped with caterpillar, has unique steering characteristics and is quite different from that of a wheeled vehicle. The steering of tracked vehicles is operated in the power pack due to different speeds of both sprockets. Under cornering conditions, power split and power regeneration are happened in the power pack. In case of power regeneration, power is transferred outside track after adding engine power and power inputted inside track from the ground. However, excessive power regeneration is transferred in the power pack. It damages mechanical elements. Therefore, it is necessary to analyze the steering system and check mentioned problem above. In this study, the detailed dynamic model of steering system is presented, which includes slippage between track and roadwheel, inertia force, and inertia moment. Finally, our model is compared with the Kitano model and we verified the validity of the model.

• Proceedings of the KSR Conference
• /
• 2011.05a
• /
• pp.585-593
• /
• 2011

Actual fire test under a laboratory and fire simulation by using computer are considered into main methodology in order to estimate and predict fire size of railway train. Even if practical fire size could be obtained from the full-model railway car test such as a large scale cone-calorimeter test, it is not always possible and realistic due to that expensive cost and attendant dangers could in no way be negligible. In this point of view, fire simulation analysis method based on the computational fluid dynamics could be proposed as an alternative and it seems to be also efficient and reasonable. However, simulation results have to be verified and validated in accordance with the proper procedure including comparing analysis with the actual fire test. In this paper, fire load and growth aspect was investigated through the room corner test (ISO 9705) for the mock-up model of the actual railway car. Then, it was compared with the output data derived from the simulation by using Pyrolysis Model of the FDS (Fire Dynamics Simulator, by NIST) for the exact same domain and condition corresponding with pre-performed room-corner test. This preliminary verified and validated fire modeling method could enhance the reliability of output data derived from the fire simulation under the similar domain and condition.

• Proceedings of the KSR Conference
• /
• 2011.10a
• /
• pp.2852-2859
• /
• 2011

Planning and construction of railway for high speed trains up to 400 km/h are recently driven in Korea. High speed train is one of the environment-friendly fastest mass transportation means but its noise generated by rolling, traction and aerodynamic mechanism can cause public complaints of residents nearby railways. To cost-effectively prevent the troublesome noise in a railway planning stage, the rational railway noise prediction method considering the characteristics of trains as well as railway structures should be required but it is difficult to find an authentic method for Korean high speed trains such as KTX and KTX-II. In this study, recent railway noise prediction methods developed by EU countries are introduced and discussed for consulting before setting the framework of our own railway noise prediction model emitted by Korean high speed trains over 250 km/h. Especially, the new Schall 03 model (2006) developed by Germany and IMAGINE model (2007) suggested by an EU framework research project are intensively reviewed. In addition, research items required for the development of our own model are suggested.

• Journal of Digital Contents Society
• /
• v.12 no.4
• /
• pp.467-474
• /
• 2011

Auditory and language train for the hearing-impaired is an essential course improving conversational capability with non-deaf and accompanying the financial burden and the physical fatigue of parents or a teacher. To reduce these problems, web-based training contents have been developed. But these contents have been developed without consideration of individual difference such as various levels of residual hearing and the learning capability of hearing-impaired. Therefore, it is important that appropriate training progress for each hearing-impaired should be designed by evaluating and analyzing the personal status, residual hearing, learning capability and training achievement. This paper suggests auditory and language training service model for the hearing-impaired, which is planning and managing an auditory and learning training based on personal evaluation. In addition, this paper suggests a design method for a serious game content planing based on this service model.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.10 no.5
• /
• pp.645-652
• /
• 2015

In underwater, such as fish farm and sea, turbidity is increased by water droplets and various suspended, therefore light attenuation occurs depending on the depth also caused by the scattering effect of light float. In this paper, in order to improve the visibility of underwater images obtained from these aquatic environment, we propose a visibility enhancement method using a haze removal method based on dark channel prior and a trained color transform model. In order to train a color transform model, we used underwater pattern images captured from Pohang and Yeosu, and to measure the performance of the proposed method, we carried out experiment of visibility enhancement using underwater images collected from Yeosu, Geomundo and Philippines. The results show that the proposed method can improve the visibility of underwater images of various locations.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.59 no.3
• /
• pp.492-498
• /
• 2010

Since it is almost impossible to train the dispatchers with real power system, the dispatcher training simulator(DTS) is used for the training. Among various components of the DTS, the power system model(PSM) emulates the dynamic behavior of the power system to calculate the frequency and voltage. The frequency is calculated from various parameters such as mechanical power of power plants, load, inertia, and the damping of the power system. In the PSM, the power plants are modeled as differential equations, so the mechanical power of the power plants are calculated by the numerical methods. Conventionally, the fixed step-size algorithm has been used in the PSM, however it has some drawbacks. This paper develops the prototype PSM using the Matlab, and analyzes the problems of the fixed step-size algorithm by comparing the results with those of PSCAD simulation. In order to overcome the limitations, this paper proposes a modified frequency calculation method using the adaptive step-size algorithm. From the simulation using the proposed method, it is verified that the accuracy of frequency calculation increases substantially while the simulation time is not greatly increased.

• Journal of Korea Multimedia Society
• /
• v.22 no.12
• /
• pp.1447-1456
• /
• 2019

As artificial intelligence technology is actively used in image processing, it is possible to generate high-quality fake images based on deep learning. Fake images generated using GAN(Generative Adversarial Network), one of unsupervised learning algorithms, have reached levels that are hard to discriminate from the naked eye. Detecting these fake images is required as they can be abused for crimes such as illegal content production, identity fraud and defamation. In this paper, we develop a deep-learning model based on CNN(Convolutional Neural Network) for the detection of StyleGAN fake images. StyleGAN is one of GAN algorithms and has an excellent performance in generating face images. We experiment with 48 number of experimental scenarios developed by combining parameters of the proposed model. We train and test each scenario with 300,000 number of real and fake face images in order to present a model parameter that improves performance in the detection of fake faces.