• The Korean Journal of Applied Statistics
• /
• v.11 no.2
• /
• pp.415-421
• /
• 1998

In this paper, we proposed improved unrelated question model which has the benefit of simplicity the Kim et al.'s two-stage unrelated question model(1992). conditions are obtained under which the proposed model is more efficient than the Greenberg et al. model(1971) and Kim et al's two-stage unrelated question model.

• Communications for Statistical Applications and Methods
• /
• v.30 no.5
• /
• pp.501-516
• /
• 2023

Knowledge distillation (KD) refers to extracting knowledge from a large and complex model (teacher) and transferring it to a relatively small model (student). This can be done by training the teacher model to obtain the activation function values of the hidden or the output layers and then retraining the student model using the same training data with the obtained values. Recently, relational KD (RKD) has been proposed to extract knowledge about relative differences in training data. This method improved the performance of the student model compared to conventional KDs. In this paper, we propose a new method for RKD by introducing a new loss function for RKD. The proposed loss function is defined using the area difference between the teacher model and the student model in a specific hidden layer, and it is shown that the model can be successfully compressed, and the generalization performance of the model can be improved. We demonstrate that the accuracy of the model applying the method proposed in the study of model compression of audio data is up to 1.8% higher than that of the existing method. For the study of model generalization, we demonstrate that the model has up to 0.5% better performance in accuracy when introducing the RKD method to self-KD using image data.

• The KIPS Transactions:PartB
• /
• v.17B no.1
• /
• pp.37-46
• /
• 2010

In this paper we have presented a two dimensional model based tracking system using improved chamfer matching. Conventional chamfer matching could not calculate similarity well between the object and image when there is very cluttered background. Then we have improved chamfer matching to calculate similarity well even in very cluttered background with edge and corner feature points. Improved chamfer matching is used as likelihood function of particle filter which tracks the geometric object. Geometric model which uses edge and corner feature points, is a discriminant descriptor in color changes. Particle Filter is more non-linear tracking system than Kalman Filter. Then the presented method uses geometric model, particle filter and improved chamfer matching for tracking object in complex environment. In experimental result, the robustness of our system is proved by comparing other methods.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.11 no.1
• /
• pp.79-89
• /
• 2006

Particularly the photovoltaic systems are preferred because the output is extracted to the useful electric energy. However, the output characteristics of photovoltaic(PV) systems using solar cell or array depend on the weather conditions. The assistant equipment which emulates the solar cell characteristics that can be controlled arbitrarily by researcher is required to the researchers for reliable experimental data. To solve these problems, it is necessary to research a solar cell model of which output characteristics varied by setting the weather conditions such as insolation levels and temperatures. Therefore, this paper was presented that improved model which is based on interpolation model. To verified the improved model, it is confirmed using the simulation of MATLAB. Also, the experiment was performed by the characteristics of virtual implemented solar cell(VISC) system with the proposed solar cell model. It could be confirmed that there exists actual ewer within 5% between actual solar cell and VISC system.

• Geomechanics and Engineering
• /
• v.34 no.5
• /
• pp.491-505
• /
• 2023

The practical usage of underground space and demand for vehicular tunnels necessitate the construction of non-circular wide rectangular tunnels. However, constructing large tunnels in soft clayey soil conditions with no ground improvement can lead to excessive ground deformations and collapse. In recent years, in situ ground improvement techniques such as jet grouting and deep cement mixing are often utilized to perform cement-stabilisation around the tunnel boundary to prevent large deformations and failure. This paper discusses the stability characteristics and failure behaviour of a wide rectangular tunnel in cement-treated soft clays. First, the plane strain finite element model is developed and validated with the results of centrifuge model tests available in the past literature. The critical tunnel support pressures computed from the numerical study are found to be in good agreement with those of centrifuge model tests. The influence of varying strength and thickness of improved soil surround, and cover depth are studied on the stability and failure modes of a rectangular tunnel. It is observed that the failure behaviour of the tunnel in improved soil surround depends on the ratio of the strength of improved soil surround to the strength of surrounding soil, i.e., q_ui/q_us, rather than just q_ui. For low q_ui/q_us ratios,the stability increases with the cover; however, for the high strength improved soil surrounds with q_ui >> q_us, the stability decreases with the cover. The failure chart, modified stability equation, and stability chart are also proposed as preliminary design guidelines for constructing rectangular tunnels in the improved soil surrounded by soft clays.

• Smart Structures and Systems
• /
• v.33 no.5
• /
• pp.365-374
• /
• 2024

This paper adopts a new approach in which nonlinear vibrations can be controlled using fuzzy controllers by optimal grey evolutionary algorithm. If the fuzzy controller cannot stabilize the systems, then the high frequency is injected into the system to assist the controller, and the system is asymptotically stabilized by adjusting the parameters. This paper uses the GM (grey model) and the neural network prediction model. The structure of the neural network is improved from a single factor, and multiple data inputs are extended to various factors and numerous data inputs. The improved model expands the applicable range of uncontrolled elements and improves the accuracy of controlled prediction, using the model that has been trained and stabilized by multiple learning. The simulation results show that the improved gray neural network model has higher prediction accuracy and reliability than the traditional GM model, improving controlled management and pre-control ability. In the combined prediction, the time series parameters and the predicted values obtained from the GM (1,1) (Grey Model of first order and one variable) are simultaneously used as the input terms of the neural network, considering the influence of the non-equal spacing of the data, which makes the results of the combined gray neural network model more rationalized. By adjusting the model structure and system parameters to simulate and analyze the controlled elements, the corresponding risk change trend graphs and prediction numerical calculation results are obtained, which also realize the effective prediction of controlled elements. According to the controlled warning principle and objective, the fuzzy evaluation method establishes the corresponding early warning response method. The goals of this paper are towards access to adequate, safe and affordable housing and basic services, promotion of inclusive and sustainable urbanization and participation, implementation of sustainable and disaster-resilient buildings, sustainable human settlement planning and manage.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.9
• /
• pp.777-781
• /
• 2001

This paper presents an image contrast enhancement technique for improving the low contrast images using the improved IAFC(Integrated Adaptive Fuzzy Clustering) model. The low pictorial information of a low contrast image is due to the vagueness or fuzziness of the multivalued levels of brightness rather than randomness. Fuzzy image processing has three main stages, namely, image fuzzification, modification of membership values, and image defuzzification. Using a new model of automatic crossover point selection, optimal crossover point is selected automatically. The problem of crossover point selection can be considered as the two-category classification problem. The improved IAFC model is used to classify the image into two classes. The proposed method is applied to several experimental images with 256 gray levels and the results are compared with those of the histogram equalization technique. We utilized the index of fuzziness as a measure of image quality. The results show that the proposed method is better than the histogram equalization technique.

• Nuclear Engineering and Technology
• /
• v.55 no.7
• /
• pp.2628-2641
• /
• 2023

Nuclear safety is the lifeline for the development and application of nuclear energy. In severe accidents of pressurized water reactor (PWR), aerosols, as the main carrier of fission products, are suspended in the containment vessel, posing a potential threat of radioactive contamination caused by leakage into the environment. The gas-phase aerosols suspended in the containment will settle onto the wall or sump water through the natural deposition mechanism, thereby reducing atmospheric radioactivity. Aiming at the low accuracy of the aerosol model in the ISAA code, this paper improves the natural deposition model of aerosol in the containment. The aerosol dynamic shape factor was introduced to correct the natural deposition rate of non-spherical aerosols. Moreover, the gravity, Brownian diffusion, thermophoresis and diffusiophoresis deposition models were improved. In addition, ABCOVE, AHMED and LACE experiments were selected to validate and evaluate the improved ISAA code. According to the calculation results, the improved model can more accurately simulate the peak aerosol mass and respond to the influence of the containment pressure and temperature on the natural deposition rate of aerosols. At the same time, it can significantly improve the calculation accuracy of the residual mass of aerosols in the containment. The performance of improved ISAA can meet the requirements for analyzing the natural deposition behavior of aerosol in containment of advanced PWRs in severe accident. In the future, further optimization will be made to address the problems found in the current aerosol model.

• Proceedings of the KSME Conference
• /
• 2007.05b
• /
• pp.2372-2377
• /
• 2007

Characteristics of turbulent lifted flames in coflow jets with the varying initial temperature have recently been investigated about only propane case diluted by nitrogen. The investigation has firstly improved a premixed flame model and a large scale mixing model among competing theories on the stabilization mechanism of turbulent flame to be suitable for a high temperature condition. In this research, about methane with good availability to apply for a practical combustor as clean fuel, its characteristics of turbulent nonpremixed flame have been studied experimentally. The results have shown an effectiveness of the premixed flame model and the large scale mixing model considered initial temperature variation. Additionally, considering the axial distance where the mean fuel concentration falls below the stoichiometric level along the center line of the jet according to diluting nitrogen, the premixed flame model have more accurately been improved.

• Journal of Digital Convergence
• /
• v.10 no.10
• /
• pp.277-282
• /
• 2012

The speech recognition system can not quickly adapt to varied environmental noise factors that degrade the performance of recognition. In this paper, the echo noise robust HMM learning model using average estimator LMS algorithm is proposed. To be able to adapt to the changing echo noise HMM learning model consists of the recognition performance is evaluated. As a results, SNR of speech obtained by removing Changing environment noise is improved as average 3.1dB, recognition rate improved as 3.9%.