• Journal of the Korean Magnetics Society
• /
• v.27 no.2
• /
• pp.49-53
• /
• 2017

Metamaterials composed of split cut wire (SCW) on grounded polyimide film substrate have been investigated for the aim of electromagnetic wave absorbers operated in THz frequency band. Reflection loss and current density distributions are numerically simulated with variations of the SCW geometries using the commercial software. The minimum reflection loss lower than -20 dB has been identified at 5.5~6.5 THz. The simulated resonance frequency and reflection loss can be explained on the basis of the circuit theory of an inductance-capacitance (L-C) resonator. Dual-band absorption can be obtained by arrangement of two SCWs of different length on the top layer of the grounded substrate, which is due to multiple magnetic resonances by scaling of SCWs. With increasing the side spacing between SCWs, a more enhanced absorption peak is observed at the first resonance frequency that is shifted to a lower frequency.

• Journal of KIISE:Software and Applications
• /
• v.31 no.4
• /
• pp.457-466
• /
• 2004

Recommendation is to offer information which fits user's interests and tastes to provide better services and to reduce information overload. It recently draws attention upon Internet users and information providers. The collaborative filtering is one of the widely used methods for recommendation. It recommends an item to a user based on the reference users' preferences for the target item or the target user's preferences for the reference items. In this paper, we propose a neural network based collaborative filtering method. Our method builds a model by learning correlation between users or items using a multi-layer perceptron. We also investigate integration of diverse information to solve the sparsity problem and selecting the reference users or items based on similarity to improve performance. We finally demonstrate that our method outperforms the existing methods through experiments using the EachMovie data.

• Journal of Sensor Science and Technology
• /
• v.29 no.3
• /
• pp.172-179
• /
• 2020

We studied an integrated-optic biosensor configuration that operates at a wavelength of 0.63 ㎛ based on the evanescent-wave and two horizontal mode power coupling of Si₃N₄ rib-optical waveguides formed on a Si/SiO₂/Si₃N₄/SiO₂ multilayer thin films. The sensor consists of a single-mode input waveguide, followed by a two-mode section which acts as the sensing region, and a Y-branch output for separating the two output waveguides. The coupling between the two propagating modes in the sensing region produces a periodically repeated optical power exchanges along the propagation. A light power was steered from one output channel to the other due to the change in the cladding layer (bio-material) refractive index, which affected the effective refractive index (phase-shift) of two modes through evanescent-wave. Waveguide analyses based on the rib optical waveguide dimensions were performed using various numerical computational software. Sensitivity values of 12~23 and 65~165 au/RIU, respectively for the width and length of 4 ㎛, and 3841.46 and 26250 ㎛ of the two-mode region corresponding to the refractive index range 1.36~1.43 and 1.398~1.41, respectively, were obtained.

• Journal of KIISE:Software and Applications
• /
• v.34 no.7
• /
• pp.585-594
• /
• 2007

The back propagation algorithm took a long time to learn the input patterns and was difficult to train the additional or repeated learning patterns. So Aleksander proposed the binary neural network which could overcome the disadvantages of BP Network. But it had the limitation of repeated learning and was impossible to extract a generalized pattern. In this paper, we proposed a dynamic 3 dimensional Neuro System which was consisted of a learning network which was based on weightless neural network and a feedback module which could accumulate the characteristic. The proposed system was enable to train additional and repeated patterns. Also it could be produced a generalized pattern by putting a proper threshold into each learning-net's discriminator which was resulted from learning procedures. And then we reused the generalized pattern to elevate the recognition rate. In the last processing step to decide right category, we used maximum response detector. We experimented using the MNIST database of NIST and got 99.3% of right recognition rate for training data.

• Journal of KIISE:Software and Applications
• /
• v.32 no.2
• /
• pp.85-98
• /
• 2005

Recently, a lot of researches about mobile vision using Personal Digital Assistant(PDA) has been attempted. Many CPUs for PDA are integer CPUs, which have no floating-computation component. It results in slow computation of the algorithms peformed by vision system or image processing, which have much floating-computation. In this paper, in order to resolve this weakness, we propose the Client(PDA)/server(PC) architecture which is connected to each other with a wireless LAN, and we construct the system with pipelining processing using two CPUs of the Client(PDA) and the Server(PC) in image sequence. The Client(PDA) extracts tentative text regions using Edge Density(ED). The Server(PC) uses both the Multi-1.aver Perceptron(MLP)-based texture classifier and Connected Component(CC)-based filtering for a definite text extraction based on the Client(PDA)'s tentativel99-y extracted results. The proposed method leads to not only efficient text extraction by using both the MLP and the CC, but also fast running time using Client(PDA)/server(PC) architecture with the pipelining processing.

• Tribology and Lubricants
• /
• v.37 no.4
• /
• pp.144-150
• /
• 2021

The most cost-effective method of reducing abrasive wear in mechanical parts is increasing their hardness with thin hard coatings. In practice, the composite hardness of the coated substrate is more important than that of the substrate or coating. After full unloading of the load applied to an indenter, its indentation hardness evaluated based on the dent created on the test piece was almost dependent on plastic deformation of the substrate. Following the first part of this study, which proposes a new Brinell hardness test method for a coated surface, the remainder of the study is focused on practical application of the method. Indentation analyses of a rigid sphere and elastic-perfect plastic materials were performed using finite element analysis software. The maximum principal stress and plastic strain distributions as well as the dent shapes according to the substrate yield stress and coating thickness were compared. The substrate yield stress had a significant effect on the dent size, which in turn determines the Brinell hardness. In particular, plastic deformation of the substrate produced dents regardless of the state of the coating layer. The hardness increase by coating behaved differently depending on the substrate yield stress, coating thickness, and indentation load. These results are expected to be useful when evaluating the composite hardness values of various coated friction surfaces.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.10
• /
• pp.407-420
• /
• 2021

Deep neural networks are widely used to solve various problems. However, the deep neural network with a deep hidden layer frequently has a vanishing gradient or exploding gradient problem, which is a major obstacle to learning the deep neural network. In this paper, we propose a parametric activation function to alleviate the vanishing gradient problem that can be caused by nonlinear activation function. The proposed parametric activation function can be obtained by applying a parameter that can convert the scale and location of the activation function according to the characteristics of the input data, and the loss function can be minimized without limiting the derivative of the activation function through the backpropagation process. Through the XOR problem with 10 hidden layers and the MNIST classification problem with 8 hidden layers, the performance of the original nonlinear and parametric activation functions was compared, and it was confirmed that the proposed parametric activation function has superior performance in alleviating the vanishing gradient.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.7
• /
• pp.263-270
• /
• 2021

Reinforcement learning(RL) is the method to find an optimal policy through training. and it is one of popular methods for solving lifesaving and disaster response problems effectively. However, the conventional reinforcement learning method for disaster response utilizes either simple environment such as. grid and graph or a self-developed environment that are hard to verify the practical effectiveness. In this paper, we propose the design of a disaster response RL environment which utilizes the detailed property information of the disaster simulation in order to utilize the reinforcement learning method in the real world. For the RL environment, we design and build the reinforcement learning communication as well as the interface between the RL agent and the disaster simulation. Also, we apply the dimension reduction method for converting non-image feature vectors into image format which is effectively utilized with convolution layer to utilize the high-dimensional and detailed property of the disaster simulation. To verify the effectiveness of our proposed method, we conducted empirical evaluations and it shows that our proposed method outperformed conventional methods in the building fire damage.

• KIPS Transactions on Software and Data Engineering
• /
• v.8 no.1
• /
• pp.19-26
• /
• 2019

In this paper, the classification performance of learning algorithms is compared for TAB digit recognition. The TAB digits that are segmented from TAB musical notes contain TAB lines and musical symbols. The labeling method and non-linear filter are designed and applied to extract fret digits only. The shift operation of the 4 directions is applied to generate more data. The selected models are Bayesian classifier, support vector machine, prototype based learning, multi-layer perceptron, and convolutional neural network. The result shows that the mean accuracy of the Bayesian classifier is about 85.0% while that of the others reaches more than 99.0%. In addition, the convolutional neural network outperforms the others in terms of generalization and the step of the data preprocessing.

• Journal of Hydrogen and New Energy
• /
• v.29 no.6
• /
• pp.570-577
• /
• 2018

Composite pressure vessels made through filament winding are widely used in various fields. Numerous studies regarding composite pressure vessels have been conducted in the automotive industry to improve the space efficiency of trunks as well as the fuel efficiency. Compared with steel liquefied petroleum gas (LPG) vessels used in the conventional LPG vehicles, the use of type 4 composite pressure vessels has advantages in terms of reduction of the weight of vehicles. This study focused on development of type 4 composite pressure vessels that can be installed in the spare tire well. Those type 4 composite pressure vessels are designed with torispherical dome shapes instead of geodecis dome shapes because of the space limitation. To reduce deformation due to the stresses in the axial direction of the vessels, thereby securing the safety of the container, the reinforcing bar concept was applied. A structural analysis software, ABAQUS, confirmed the effect of the reinforcing bar on the axial deformation through the type 4 composite pressure vessel. As a result, the final winding angle of the composite layer was analyzed by applying $26^{\circ}/28^{\circ}/26^{\circ}/28^{\circ}/26^{\circ}/88^{\circ}$ The tensile stress was 939.2 MPa and the compressive stress was 249.3 MPa.