• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.33 no.2
• /
• pp.201-210
• /
• 2023

Deep learning-based face verificattion model show high performance and are used in many fields, but there is a possibility the user's face image may be leaked in the process of inputting the face image to the model. Althoughde-identification technology exists as a method for minimizing the exposure of face features, there is a problemin that verification performance decreases when the existing technology is applied. In this paper, after combining the face features of other person, a de-identified face image is created through StyleGAN. In addition, we propose a method of optimizingthe combining ratio of features according to the face verification model using HopSkipJumpAttack. We visualize the images generated by the proposed method to check the de-identification performance, and evaluate the ability to maintain the performance of the face verification model through experiments. That is, face verification can be performed using the de-identified image generated through the proposed method, and leakage of face personal information can be prevented.

• Journal of Information Technology Services
• /
• v.13 no.1
• /
• pp.253-267
• /
• 2014

This study explores to identify the characteristics of the business model by comparing and analyzing the value creation between two cases, and to be successful in business model innovation. In order for the pursuit of purposes, domestic and international firm' business model cases are analyzed. Regarding the business model innovation, huge differences are found between two cases. First, a clear customer value proposal is important. Netflix is constantly monitoring the customer's needs and satisfactions to improve value proposition, while Joycube, domestic firm does not adjust to meet the change of the customer's behaviors. Second, the business model innovation should be taking into account the customer's behaviors in the constant changing market environments. For the growth, firms should consider strategic monitoring the market environments, and find a novelty of the markets, and to create the jump through business model innovation.

• Journal of Korea Port Economic Association
• /
• v.34 no.1
• /
• pp.99-110
• /
• 2018

This study empirically examines simple methodology to quantify the risk resulted from the uncertainty of bunker price and foreign exchange rate, which cause main resources of the cost in shipping industry during the periods between $1^{st}$ of January 2010 and $31^{st}$ of January 2018. To shed light on the risk measurement in cash flows we tested GBM(Geometric Brownian Motion) frameworks such as the model with conditional heteroskedasticity and jump diffusion process. The main contribution based on empirical results are summarized as following three: first, the risk analysis, which is dependent on a single variable such as freight yield, is extended to analyze the effects of multiple factors such as bunker price and exchange rate return volatility. Second, at the individual firm level, the need for risk management in bunker price and exchange rate is presented as cash flow. Finally, based on the scale of the risk presented by the analysis results, the shipping companies are required that there is a need to consider what is appropriate as a means of risk management.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.24 no.5
• /
• pp.577-588
• /
• 2011

This paper provides a novel extended Moving Least Squares(MLS) difference method for the potential problem with weak and strong discontinuities. The conventional MLS difference method is enhanced with jump functions such as step function, wedge function and scissors function to model discontinuities in the solution and the derivative fields. When discretizing the governing equations, additional unknowns are not yielded because the jump functions are decided from the known interface condition. The Poisson type PDE's are discretized by the difference equations constructed on nodes. The system of equations built up by assembling the difference equations are directly solved, which is very efficient. Numerical examples show the excellence of the proposed numerical method. The method is expected to be applied to various discontinuity related problems such as crack problem, moving boundary problem and interaction problems.

• Journal of Mechanical Science and Technology
• /
• v.18 no.11
• /
• pp.1969-1977
• /
• 2004

To model and understand the physics of partial rub, a nonlinear rotor model is sought by applying a nonlinear parameter identification technique to the experimental data. The results show that the nonlinear terms of damping and stiffness should be included to model partial rotor rub. Especially, the impact and friction during the contact between rotor and stator are tried to explain with a nonlinear model on the basis of experimental data. The estimated nonlinear model shows good agreements between the numerical and the experimental results in its orbit. Also, the estimated nonlinear model could explain the backward whirling orbit and jump phenomenon, which are the typical phenomena of partial rub.

• The KSFM Journal of Fluid Machinery
• /
• v.20 no.1
• /
• pp.15-20
• /
• 2017

The small turbocharger for the automotive application is designed to operate up to 200,000 rpm to increase system efficiency. Because of high rotation speed of turbocharger, floating ring bearing are widely adopted due to its low friction loss and high rotordynamic stability. This paper presents a linear and nonlinear analysis model for a turbocharger rotor supported by a semi-floating ring bearing. The rotordynamic model for the turbocharger rotor was constructed based on the finite element method and fluid film forces were calculated based on the infinitely short bearing assumption. In linear analysis, we considered fluid film force as stiffness and damping element and in nonlinear analysis, the fluid film force was calculated by solving the time dependent Reynolds equation. We verified the developed theoretical model by comparing to modal test results of test rotors. The analysis results show that there are two unstable modes, which are conical and cylindrical modes. These unstable modes appear as sub-synchronous vibrations in nonlinear analysis. In nonlinear analysis, frequency jump phenomenon demonstrated when vibration mode is changed from conical mode to cylindrical one. This jump phenomenon was also demonstrated in the test. However, the natural frequency measured in the test differs from those obtained using nonlinear analysis.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.05a
• /
• pp.277-282
• /
• 2001

Partial rotor rub occurs when an obstacle on the stator of a rotating machinery disturbs the free whirling motion of a rotor, which is more common than full annular rub for the cases of rubbing in rotating machinery. The nonlinearity due to the intermittent contacts and friction during partial rotor rub makes the phenomenon complex. The several nonlinear phenomena of superharmonics, subharmonics, and jump phenomenon are demonstrated for the partial rub using an experimental apparatus in this study. A piecewise-linear model and a rebound model using the coefficient of restitution are investigated on the basis of experimental observations in order to adopt as an analytical model of the contact between the rotor and stator during whirling motion. The contact stiffness, coefficient of restitution, and friction coefficient for the contact during partial rub are calculated from the comparison between the numerical simulation and the experimental results. Also, the numerical simulations for the model of partial rub are done for the various system parameters of clearance, contact stiffness, and friction coefficient in order to find the nonlinear behavior of partial rotor rub.

• Journal of Environmental Science International
• /
• v.22 no.2
• /
• pp.173-185
• /
• 2013

In order to clarify the impact of wildfire and its thermal forcing on atmospheric wind and temperature patterns, several numerical experiments were carried out using three dimensional atmospheric dynamic model WRF with wildfire parametrization module SFIRE. Since wind can accelerate fire spread speed, the moving speed of fireline is faster than its initial values, and the fireline tends to move the northeast, because of the wind direction and absolute vorticity conservation law associated with driving force induced by terrain. In comparison with non-fire case, the hydraulic jump that often occurs over downwind side of mountain became weak due to huge heat flux originated by surface wildfire and wind pattern over downwind side of mountain tends to vary asymmetrically with time passing. Therefore temporal variation of wind pattern should be catched to prevent the risk of widfire.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.6
• /
• pp.459-464
• /
• 2001

This paper discusses a stochastic stabilization of Takagi-Sugeno(TS) fuzzy system with Markovian input delay. The finite Markovian process is adopted to model the input delary of the overall control system. It is assumed that the zero and hold devices are used for control input. The continuous-time TS fuzzy system with the Markovian input delay is discretized for easy handling delay, according, the discretized TS fuzzy system is represented by a discrete-time TS fuzzy system with jumping parameters. The stochastic stabilizibility of the jump TS fuzzy system is derived and formulated in terms of linear matrix inequalities (LNIS)

• Asia pacific journal of information systems
• /
• v.7 no.1
• /
• pp.67-83
• /
• 1997

The utility of learning techniques in investment analysis has been demonstrated in many areas, ranging from forecasting individual stocks to entire market indexes. To date, however, the application of artificial intelligence to financial forecasting has focused largely on short predictive horizons. Usually the forecast window is a single period ahead; if the input data involve daily observations, the forecast is for one day ahead; if monthly observations, then a month ahead; and so on. Thus far little work has been conducted on the efficacy of long-term prediction involving multiperiod forecasting. This paper examines the impact of alternative procedures for extended prediction using knowledge discovery techniques. One dimension in the study involves temporal granularity: a single jump from the present period to the end of the forecast window versus a web of short-term forecasts involving a sequence of single-period predictions. Another parameter relates to the numerosity of input variables: a technical approach involving only lagged observations of the target variable versus a fundamental approach involving multiple variables. The dual possibilities along each of the granularity and numerosity dimensions entail a total of 4 models. These models are first evaluated using neural networks, then compared against a multi-input jump model using case based reasoning. The computational models are examined in the context of forecasting the S&P 500 index.