• ETRI Journal
• /
• 제28권1호
• /
• pp.59-66
• /
• 2006

Learning behaviors of a radial basis function network (RBFN) using a singular value decomposition (SVD) and stochastic gradient (SG) algorithm, together named RBF-SVD-SG, for odor sensing systems are analyzed, and a fast training method is proposed. RBF input data is from a conducting polymer sensor array. It is revealed in this paper that the SG algorithm for the fine-tuning of centers and widths still shows ill-behaving learning results when a sufficiently small convergence coefficient is not used. Since the tuning of centers in RBFN plays a dominant role in the performance of RBFN odor sensing systems, our analysis is focused on the center-gradient variance of the RBFN-SVD-SG algorithm. We found analytically that the steadystate weight fluctuation and large values of a convergence coefficient can lead to an increase in variance of the center-gradient estimate. Based on this analysis, we propose to use the least mean square algorithm instead of SVD in adjusting the weight for stable steady-state weight behavior. Experimental results of the proposed algorithm have shown faster learning speed and better classification performance.

• 한국수자원학회논문집
• /
• 제35권2호
• /
• pp.187-194
• /
• 2002

하천유역 면적강우량 산정의 정확도를 개선하기 위하여 기존 강우관측자료의 통계적 특성을 이용한 강우관측망의 최적설계방법을 연구하였다. 최적설계를 위한 목적함수는 면적강우량의 추정오차 및 지점강우량 관측비용의 항으로 구성하고, 그 값이 최소인 관측망은 선정하였다. 통계f7파의 추정방법으로는 통계적 분산 산정방법인 크리깅 모형을 채택하였다. 비용은 강우관측소의 설치비와 연간운영 비론 적용하고, 오차항과 비용항의 통합에는 등치매개변수를 이용하였다. 연구된 최적설계방법을 댐 신설로 강우관측소 증설이 필요한 용담댐 유역에 적용하여, 대상유역의 최적 강우관측망을 제안하였다.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제26권4호
• /
• pp.296-309
• /
• 2022

Timer options are one of the contingent claims that, for given the variance budget, its payoff depends on a random maturity in terms of the realized variance unlike the standard European vanilla option with a fixed time maturity. Since it was first launched by Société Générale Corporate and Investment Banking in 2007, the valuation of the timer options under several stochastic environment for the volatility has been conducted by many researches. In this study, we propose the pricing of timer power options combined with standard timer options and the index of the power to the underlying asset for the investors to actualize lower risks and higher returns at the same time under the uncertain markets. By using the asymptotic analysis, we obtain the first-order approximation of timer power options. Moreover, we demonstrate that our solution has been derived accurately by comparing it with the solution from the Monte-Carlo method. Finally, we analyze the impact of the stochastic volatility with regards to various parameters on the timer power options numerically.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.172-175
• /
• 2002

This paper presents perceptual model with a stochastic rnultiresolution characteristic that can be applied with watermark embedding in the biorthogonal wavelet domain. The perceptual model with adaptive watermarking algorithm embed at the texture and edge region for more strongly embedded watermark by the SSQ(successive subband quantization). The watermark embedding is based on the computation of a NVF(noise visibility function) that have local image properties. This method uses non-stationary Gaussian model stationary Generalized Gaussian model because watermark has noise properties. In order to determine the optimal NVF, we consider the watermark as noise. The particularities of embedding in the stationary GG model use shape parameter and variance of each subband regions in multiresolution. To estimate the shape parameter, we use a moment matching method. Non-stationary Gaussian model use the local mean and variance of each subband. The experiment results of simulation were found to be excellent invisibility and robustness. Experiments of such distortion are executed by Stirmark benchmark test.

• 한국환경보건학회지
• /
• 제17권2호
• /
• pp.1-8
• /
• 1991

The stochastic variance and structures of time series data on air quality were examined by employing the techniques of autocorrelation function, variance spectrum, fourier series, ARIMA model. Among the air quality properties of atmosphere, SO$_{2}$ is one of the most siginificant and widely measured parameters. In the study, the air quality data were included hourly observations on SO$_{2}$ TSP and O$_{3}$. The data were measured by automatic recording instrument installed in Kwanghwamoon during February and March in 1991. The results of study were as follows 1. Hourly air quality series varied with the domiant 24 hour periodicity and the 12 hour periodic variation was also observed. 2. The correlation coefficients between SO$_{2}$ and O$_{3}$ is -0.4735. 3. In simulating or forecasting variation in SO$_{2}$ ARIMA models are on a useful tools. The multiplicative seasonal ARIMA (1, 1, 0) (0, 2, 1)$_{24}$ model provided satisfactory results for hourly SO$_{2}$ time series.

• 한국해양공학회지
• /
• 제8권2호
• /
• pp.47-55
• /
• 1994

The purpose of the present study is to investigate the effect of specimen thickness on statistical properties of crack growth resistance. In this study, the resistance S$\delta$$_h$(x) to fatigue crack growth was treated as a spatial stochastic process. which varies randomly on the crack surface. The theoretical autocorrelation functions of the resistance to fatigue crack growth considering specimen thickness are discussed for several correlation lengths. The main results obtained are : (1) The theoretical autocorrelation functions of S$\delta$$_h$(x) are almost independent of specimen of specimen thickness except for the origin. (2) The variance increases with decreasing specimen thickness.

• 한국수산과학회지
• /
• 제20권5호
• /
• pp.457-462
• /
• 1987

Stochastic stability criterias for ocean wave analysis add simulation are studied using the data simulated by the linear superposition method. To clarify the criterias, the effects of the simulation parameters on the variance of stochastic properties of ocean waves are investigated, and the stable conditions of the parameters are estimated through the comparative study on the stochastic properties of simulated waves and well-known ocean waves. The simulation parameters considered are high frequency cut-off, data length, and number and phase angle of component waves. Statistical characteristics analysed are wave height, period and steepness, and the formation of groups of higher waves, resonance periods, steeper higher waves and extreme run-length of the run.

• Structural Engineering and Mechanics
• /
• 제3권3호
• /
• pp.273-287
• /
• 1995

Stochastic response of systems to random excitation can be estimated by direct integration methods in the time domain such as the stochastic central difference method (SCDM). In this paper, the SCDM is applied to compute the variance and covariance in response of linear and nonlinear structures subjected to random excitation. The accuracy of the SCDM is assessed using two-DOF systems with both deterministic and random material properties excited by white noise. For the former case, closed-form solutions can be obtained. Numerical results also are presented for a simply supported geometrically nonlinear beam. The stiffness of this beam is modeled as a random field, and the beam is idealized by the stochastic finite element method. A perturbation technique is applied to formulate the equations of motion of the system, and the dynamic structural response statistics are obtained in a time domain analysis. The effect of variations in structural parameters and the numerical stability of the SCDM also are examined.

• 한국경영과학회지
• /
• 제20권1호
• /
• pp.179-193
• /
• 1995

A multi-objective, single-stage scheduling problem with stochastic processing times is considered where the objective is to simultaneously minimize the expected value and the variance of total flowtime, and the mean probability of tardiness. In cases where processing times follow normal distributions, a method using pairwise interchange of two jobs(PITJ) is proposed to generate a set of the approximate efficient schedules. The efficient schedules are not dominated by the criterion vectors of any other permutation schdules in the feasible region. Numerical experiments performed to ascertain the effectiveness of PITJ algorithm are also reported in the results.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1989년도 추계학술대회 논문집 학회본부
• /
• pp.191-193
• /
• 1989

This paper presents a optimal Var allocation algorithm for minimizing power loss and improving voltage profile in a given system. In this paper, nodal input data is considered as Gaussian distribution with their mean value and their variance. A stochastic Linear Programming technique based on chance constrained method is applied to solve the probabilistic constraint. The test result in IEEE-14 Bus model system showes that the voltage distribution of load buses is improved and the power loss is more reduced than before Var allocation.