• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.67 no.1
• /
• pp.114-123
• /
• 2018

This paper concerns a design and learning method of softmax function neural networks based on K-means clustering. The partial discharge data Information is preliminarily processed through simulation using an Epoxy Mica Coupling sensor and an internal Phase Resolved Partial Discharge Analysis algorithm. The obtained information is processed according to the characteristics of the pattern using a Motor Insulation Monitoring System program. At this time, the processed data are total 4 types that void discharge, corona discharge, surface discharge and slot discharge. The partial discharge data with high dimensional input variables are secondarily processed by principal component analysis method and reduced with keeping the characteristics of pattern as low dimensional input variables. And therefore, the pattern classifier processing speed exhibits improved effects. In addition, in the process of extracting the partial discharge data through the MIMS program, the magnitude of amplitude is divided into the maximum value and the average value, and two pattern characteristics are set and compared and analyzed. In the first half of the proposed partial discharge pattern classifier, the input and hidden layers are classified by using the K-means clustering method and the output of the hidden layer is obtained. In the latter part, the cross entropy error function is used for parameter learning between the hidden layer and the output layer. The final output layer is output as a normalized probability value between 0 and 1 using the softmax function. The advantage of using the softmax function is that it allows access and application of multiple class problems and stochastic interpretation. First of all, there is an advantage that one output value affects the remaining output value and its accompanying learning is accelerated. Also, to solve the overfitting problem, L2-normalization is applied. To prove the superiority of the proposed pattern classifier, we compare and analyze the classification rate with conventional radial basis function neural networks.

• Journal of Nuclear Fuel Cycle and Waste Technology(JNFCWT)
• /
• v.10 no.3
• /
• pp.161-170
• /
• 2012

To provide domestic values of input parameters in a safety assessment of radioactive waste disposal under domestic deep underground environments, various kinds of experiments have been carried out under KURT (KAERI Underground Research Tunnel) conditions. The input parameters were classified, and some of them were selected for this study by the criteria of importance. The domestic experimental data under KURT environments were given top priority in the data review process. Foreign data under similar conditions to KURT were also gathered. The collected data were arranged and the statistical calculations were processed. The properties and distribution of the data were explained and compared to foreign values in view of their validity. The following parameters were analysed: failure time and early time failure rate of a container, solubility of nuclides, porosity and density of the buffer, and distribution coefficients of nuclides in the geomedia, hydraulic conductivity, diffusion depth of nuclides, groundwater flow rate, fracture aperture, length of internal fracture, and width of faulted rock mass in the host rock.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.1
• /
• pp.56-62
• /
• 2018

In this study, an artificially formed Gap graded soil, designed to increase its shear strength, was analyzed to determine the strength parameters through direct shear tests. Uniform and fine grain size samples were compared to the Gap graded soil to investigate the increase in the shear strength. Plate loading tests were conducted using 13mm and 19mm aggregates to confirm the reproducibility of the strength enhanced samples for site application. This test confirmed that the particle size ratio and the internal friction angle are correlated to the shear strength, and the shear resistance angle significantly increased in the specific particle size ratio range. The calculation of the ultimate bearing capacity by the plate load test demonstrated that the grain size adjustment method greatly influences the strength increase rate. Therefore, the findings were verified and it was confirmed that a high shear strength is achievable despite the existence of a poor particle size distribution.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.6
• /
• pp.2575-2587
• /
• 2011

Whereas the early financial information technology invested a large amount of investment on developing and adopting its own new systems. it is now focusing on reducing expenses and improving internal efficiency but applied for securing a company's competitive advantage considering the long term strategic level. As relevant researches, studies on information system quality and performance are actively going on. In previous research, the relations of the influence on the financial performance by information system quality or absorption capacity have been reviewed. In other words, the information system quality and absorptive capacity, respectively, in the position of the independent variables have been studied as a factor to have a significant effect on financial performance while the research on the relationship between these two variables is lacking. Thus, in this researches, quality of financial information systems and the absorptive capacity, respectively, as independent variables and parameters on the mediating effect on the financial performance were researches. As a result, quality of financial information systems showed a positive effect on the performance of company, the absorption capacity was perfect to play the role of mediating one. However, absorptive capacity as an independent variable has a positive impact on financial performance, but the quality of financial information systems as a parameter was not affected in any manner.

• Journal of the Microelectronics and Packaging Society
• /
• v.15 no.1
• /
• pp.19-25
• /
• 2008

In this paper, the front end module (FEM) was proposed for 2.4GHz WLAN band by LTCC multilayer application. The FEM was composed of power amplifier IC, switch IC, and LTCC module. LTCC module consists of output matching circuit and lowpass filter as Tx part, bandpass filter as Rx part. Design of output matching circuit for LTCC was used matching parameter from output matching circuit based on lumped circuit on the PCB board. The dielectric constant of LTCC substrate is 9. The substrate was composed of total 26 layers with each 30um thickness. Ag paste was used for the internal pattern as the conductor material. The size of the module is $4.5mm{\times}3.2mm{\times}1.4mm$. The fabricated FEM showed the gain of 21dB, ACPR of less than -31dBc first side lobe and Less than -59dBc second side lobe and the output power of 23Bm at P1dB.

• Investigative Magnetic Resonance Imaging
• /
• v.21 no.1
• /
• pp.9-19
• /
• 2017

Background: Normalized cerebral blood volume (nCBV) can be measured using manual or semiautomatic segmentation method. However, the difference in diagnostic performance on brain tumor differentiation between differently measured nCBV has not been evaluated. Purpose: To compare the diagnostic performance of manually obtained nCBV to that of semiautomatically obtained nCBV on glioblastoma (GBM) and primary central nervous system lymphoma (PCNSL) differentiation. Materials and Methods: Histopathologically confirmed forty GBM and eleven PCNSL patients underwent 3T MR imaging with dynamic susceptibility contrast-enhanced perfusion MR imaging before any treatment or biopsy. Based on the contrast-enhanced T1-weighted imaging, the mean nCBV (mCBV) was measured using the manual method (manual mCBV), random regions of interest (ROIs) placement by the observer, or the semiautomatic segmentation method (semiautomatic mCBV). The volume of enhancing portion of the tumor was also measured during semiautomatic segmentation process. T-test, ROC curve analysis, Fisher's exact test and multivariate regression analysis were performed to compare the value and evaluate the diagnostic performance of each parameter. Results: GBM showed a higher enhancing volume (P = 0.0307), a higher manual mCBV (P = 0.018) and a higher semiautomatic mCBV (P = 0.0111) than that of the PCNSL. Semiautomatic mCBV had the highest value (0.815) for the area under the curve (AUC), however, the AUCs of the three parameters were not significantly different from each other. The semiautomatic mCBV was the best independent predictor for the GBM and PCNSL differential diagnosis according to the stepwise multiple regression analysis. Conclusion: We found that the semiautomatic mCBV could be a better predictor than the manual mCBV for the GBM and PCNSL differentiation. We believe that the semiautomatic segmentation method can contribute to the advancement of perfusion based brain tumor evaluation.

• Journal of Distribution Science
• /
• v.13 no.6
• /
• pp.71-78
• /
• 2015

Purpose - This study attempts to analyze whether or not there are characteristics among the top management of companies that promote corporate performance at venture companies. It investigates the characteristics of the human and social capital that are inherent in top management at a venture company and conducts an empirical analysis of hypotheses examining if these characteristics will affect utilization of the governmental R&D support system as well as affect the firm's initial success. Research design, data, and methodology - This study conducted theoretical and empirical research together to accomplish the goal of the study. The pilot study researched human capital and social capital as the independent variables; the governmental R&D support system as the parameter; and, the initial success as the dependent variable. The empirical study carried out research on the model, establishment of hypotheses, and the statistical treatment. A survey was conducted targeting top management of high-tech venture companies in Daedeok Innopolis; 500 questionnaires were distributed; and, 222 were collected. Results - The human and social capital inherent in top management at venture companies in the early stages of their existence become good evaluation data for those who are invested in similar resources. If top management includes strong human and social capital, access to external resources will be easier; these will have a positive influence on the selection of overnmental support systems; and, this proper support will also have a positive influence on the initial success of the venture company. The results revealed the following. First, it was found that when the educational level and functional background, (the top management human capital), are the output function, top management human capital had a significant influence on selection of governmental R&D support funds. Second, it was found that the internal social capital and external social capital, (the top management social capital), had a significant influence on selection of governmental R&D support tasks. Third, it was found that selection of the governmental R&D support tasks at the start of the venture company had a positive influence on the corporate financial performance such as sales, business profits, and the increase in workers; and, had a significant influence on nonfinancial performance such as market share, competitive position, product competitiveness, and the future product development. Conclusions - Selection of the governmental R&D support system is not recognized as part of the direct sales of a venture company in its early stages, but as it can reduce costs for technical development and helps significantly in creating test products and mass production, it has a positive influence on the company's financial performance and nonfinancial performance as a result. Therefore, companies should take great efforts to frequently be selected as a candidate in the governmental R&D support system, as it can help facilitate R&D that requires extensive funds. As a result, companies can expect effects such as job creation and patent applications and they can advance future product sales.

• The Korean Journal of Nuclear Medicine
• /
• v.29 no.1
• /
• pp.48-53
• /
• 1995

This study investigates a simple mathematical model for the quantitative estimation of regional myocardial blood flow in experimental canine coronary artery thrombosis using Rb-82 dynamic myocardial positron emission tomography. The coronary thrombosis was induced using the new catheter technique by narrowing the lumen of coronary vessel gradually, which finally led to partial obstruction of coronary artery. Ten Rb-82 dynamic myocardial PET scans were performed sequentially for each experiment using our 5, 10 and 20 second acquisition protocol, respectively, and three regions of interest were drawn on the transaxial slices, one on left ventricular chamber for input function and the other two on normal and decreased perfusion segments for the flow estimation in those regions. Single compartment model has been applied to the measured sets of regional PET data, and the rate constants of influx to myocardial tissue were calculated for regional myocardial flow estimates with the three parameter fits of raw data by the Levenberg-Marquardt method. The results showed that, (1) single compartment model suggested by Kety-Schmidt could be used for the simple estimation of regional myocardial blood flow, (2) the calculated regional myocardial blood flow estimates were dependent on the selection of input function, which reflected partial volume effect and left ventricular wall motion, and (3) mathematically fitted input and tissue time activity curves were more suitable than the direct application of the measured data in terms of convergence.

• Journal of the Korean GEO-environmental Society
• /
• v.10 no.6
• /
• pp.19-25
• /
• 2009

The current study developed light-weighted mixed soil that can solve problems related with soft soil such as ground subsidence, sliding and lateral displacement of ground. By reducing weight of reclaimed soil through mixing phosphogypsum and recycled EPS beads with the weathered granite soil. A series of geotechnical laboratory tests including physical index test, compaction test, CBR test, and direct shear test were performed and engineering properties were reviewed in order to assess applicability of the light-weighted mixed soil for roads and abutment and various back-filling materials at the reclamation area. Based on the laboratory test results, it was found that the maximum dry unit weight of the light-weighted soil ranges $14.32{\sim}15.79kN/m^3$ and the optimum water content ranges 21.91~24.23%, which means there is 11~19.3% weight decrease effect when comparing with general weathered granite soil. Also it was found that the corrected CBR value ranges 10.4~18.4% satisfying the domestic regulations on road subgrade and back-filling material. In addition, as for shear strength parameter, cohesion ranges 10.79~18.64 kPa and internal frictional angle ranges $35.4{\sim}37.2^{\circ}$, which are similar with those of general construction soil and back-filling material used in Korea. So it can be concluded that light-weighted mixed soil with phosphogypsum can be used effectively for soft reclamation ground as actual filling material and back-filling material. From the current study, it was found that light-weighted mixed soil with phosphogypsum has not only weight reduction effect, but also has no special problems in shear strength and bearing capacity. Therefore, it is expected that phosphogypsum can be recycled in bulk as road subgrade and back-filling material at the reclamation area.

• Journal of the Korean GEO-environmental Society
• /
• v.14 no.9
• /
• pp.31-37
• /
• 2013

In this study installation diameter, interval, area replacement ratio and ground hardness of applicable ground in C.G.S method should be mastered through surrounding ground by conducting modeling. Optimum artificial neural network was selected through the study of the parameter of artificial neural network and prediction model was developed by the relationship with numerical analysis and artificial neural network. As this result, C.G.S pile settlement and ground settlement were found to be equal in terms of diameter, interval, area replacement ratio and ground hardness, presented in a single curve, which means that the behavior pattern of applied ground in C.G.S method was presented as some form, and based on such a result, learning the artificial neural network for 3D behavior was found to be possible. As the study results of artificial neural network internal factor, when using the number of neural in hidden layer 10, momentum constant 0.2 and learning rate 0.2, relationship between input and output was expressed properly. As a result of evaluating the ground behavior of C.G.S method which was applied to using such optimum structure of artificial neural network model, is that determination coefficient in case of C.G.S pile settlement was 0.8737, in case of ground settlement was 0.7339 and in case of ground heaving was 0.7212, sufficient reliability was known.