• 품질경영학회지
• /
• 제33권1호
• /
• pp.73-82
• /
• 2005

This article proposes a practical approach, which is based on the concept of the expected squared relative error, that can consider both the prediction quality and the practitioner's subjectivity in simultaneously optimizing multiple responses. Through a case study, multiresponse optimization using the expected squared relative error approach is illustrated, and the SAS program to implement the proposed method is provided.

• 산업경영시스템학회지
• /
• 제20권41호
• /
• pp.213-220
• /
• 1997

Process capability Indices compare the actual performance of manufacturing process to the desired performance. The relationship between the capability index Cpm and the expected squared error loss provides an intuitive interpretation of Cpm. By putting the loss in relative terms a user needs only to specify the target and the distance from the target at which the product would have zero worth, or alternatively, the loss at the specification limits. Confidence limits for the expected relative loss are discussed, and numerical illustration is given.

• 한국습지학회지
• /
• 제21권spc호
• /
• pp.51-60
• /
• 2019

본 연구에서는 기후변화 시나리오 자료를 이용하여 지점빈도해석(At-site Frequency Analysis, AFA)과 지역빈도해석(Regional Frequency Analysis, RFA) 등을 수행하였고, Monte Carlo simulation을 통한 RRMSE(relative root mean squared error) 값을 비교·분석함으로써 각 빈도해석 방법에 따른 성능을 평가하고자 하였다. 확률강우량 산정을 위하여 기상청에서 국가표준시나리오로 제공하는 RCM(Regional Climate Model) 자료 중 하나인 HadGEM3-RA(12.5km) 기후모델 자료로부터 우리나라 615개 지점에 대한 일 강우 자료를 추출하였고, 자료의 편의보정(bias correction)과 공간상세화(spatial disaggregation)를 위하여 분위사상법(quantile mapping)과 역거리제곱법(inverse distance squared method)을 적용하였다. 분석 결과 지역빈도해석 방법이 지점빈도해석보다 정확하게 확률강우량을 산정하는 것으로 나타났으며, 이는 기후변화 시나리오 기반의 확률강우량 산정시 지역빈도해석의 결과가 보다 합리적인 전망 결과를 도출할 것으로 판단된다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2001년도 제14회 신호처리 합동 학술대회 논문집
• /
• pp.915-918
• /
• 2001

In this paper we study a new convergence behavior of the least mean fourth (LMF) algorithm where the error raised to the power of four is minimized for a multiple sinusoidal input and Gaussian measurement noise. Here we newly obtain the convergence equation for the sum of the mean of the squared weight errors, which indicates that the transient behavior can differ depending on the relative sizes of the Gaussian noise and the convergence constant. It should be noted that no similar results can be expected from the previous analysis by Walach and Widrow.

• 한국통신학회논문지
• /
• 제26권12A호
• /
• pp.2043-2049
• /
• 2001

In this paper we study the convergence behavior of the least mean fourth(LMF) algorithm where the error raised to the power of four is minimized for a multiple sinusoidal input and Gaussian measurement noise. Here we newly obtain the convergence equation for the sum of the mean of the squared weight errors, which indicates that the transient behavior can differ depending on the relative sizes of the Gaussian noise and the convergence constant. It should be noted that no similar results can be expected from the previous analysis by Walach add Widrow.

• Journal of the Korean Data and Information Science Society
• /
• 제20권5호
• /
• pp.887-894
• /
• 2009

The two-stage minimum risk point estimation of mean, the probability of success in a sequence of Bernoulli trials, is considered for the case where loss is taken to be symmetrized relative squared error of estimation, plus a fixed cost per observation. First order asymptotic expansions are obtained for large sample properties of two-stage procedure. Monte Carlo simulation is carried out to obtain the expected sample size that minimizes the risk and to examine its finite sample behavior.

• The Journal of the Acoustical Society of Korea
• /
• 제18권2E호
• /
• pp.3-9
• /
• 1999

In this paper we study the convergence behavior of the least mean fourth(LMF) algorithm where the error raised to the power of four is minimized for a multiple sinusoidal input and Gaussian measurement noise. Here we newly obtain the convergence equation for the sum of the mean of the squared weight errors, which indicates that the transient behavior can differ depending on the relative sizes of the Gaussian noise and the convergence constant. It should be noted that no similar results can be expected from the previous analysis by Walach and Widrow/sup [1]/.

• International Journal of Reliability and Applications
• /
• 제12권1호
• /
• pp.61-77
• /
• 2011

One of the most important problems in the estimation of the parameter of the failure model, is the cost of experimental sampling units, which can be reduced by using any prior information available about ${\theta}$, and devising a two-stage pooling shrunken estimation procedure. We have proposed an estimator of the reliability function (R(t)) of the exponential model using two-stage time censored data when a prior value about the unknown parameter (${\theta}$) is available from the past. To compare the performance of the proposed estimator with the classical estimator, computer intensive calculations for bias, mean squared error, relative efficiency, expected sample size and percentage of the overall sample size saved expressions, were done for varying the constants involved in the proposed estimator (${\tilde{R}}$(t)).

• 대한원격탐사학회지
• /
• 제37권5_1호
• /
• pp.1043-1059
• /
• 2021

식량자원의 확보와 환경생태계에 매우 중요한 요소인 산림과 농경지는 정기적인 모니터링이 요구된다. 농림 위성 영상 자료는 우리나라의 기존 산림 및 농경지 모니터링 방법의 보완재로 효과적으로 활용될 것으로 예상되고 있다. 농림위성의 발사 이전에 사전연구로써 목표 수직기하정확도 산정을 위해 수치표고모델 분석을 수행하였다. 특히 농림위성의 주요 관심 지역인 우리나라 산악지역과 농경지의 특성을 고려하여, 경사도와 식생에 따른 분석을 수행하였다. 공주, 제주 그리고 삼척 지역에 대하여 Shuttle radar topography mission digital elevation model과 Copernicus digital elevation model를 가을과 겨울에 촬영한 드론 LiDAR 수치표면모델 그리고 국토지리정보원 5 m 수치지형모델을 기준으로 평균 상대오차를 비교했다. 그 결과 Shuttle radar topography mission digital elevation model은 8.35, 8.19, 그리고 7.49 m의 상대오차를 나타냈으며, Copernicus digital elevation model는 각각 5.65, 6.73, 그리고 7.39 m의 상대 고도 오차를 나타냈다. 남한 전체에 대하여 Shuttle radar topography mission digital elevation model과 Copernicus digital elevation model를 국토지리정보원 5 m 수치지형모델을 기준으로 지형 경사에 따른 상대 고도 오차를 나타냈다. 그 결과 경사도 0°~5° 사이에서 Shuttle radar topography mission digital elevation model과 Copernicus digital elevation model의 상대 고도 오차는 각각 약 3.62와 2.52 m로 Shuttle radar topography mission digital elevation model의 오차가 더 큰 것으로 산출되었다. 하지만 경사도가 증가함에 따라 이러한 양상은 반전되어 경사도 35° 이상에서는 각각 10.16, 그리고 11.62 m 로 Copernicus digital elevation model의 상대오차가 더 크게 나타났다.

• 한국항해학회지
• /
• 제5권2호
• /
• pp.49-58
• /
• 1981

One of the main purposes of the marine gyrocompass follow-up system is to preserve the sensitive part from the wandering error due to the frictional or torsional torque around the vertical axis. This error can be diminished through the rapid follow-up action, which minimizes the relative azimuthal angular displacement between the sensitive and follow-up parts and shortens the duration of the same displacement. But an excessive rapidity of the follow-up action would result in a sustained oscillation to the system. Therefore, to design a new type of the follow-up system, the theoretical annlysis of the problems concerned should be studied systematically by introducing the control theory. This paper suggest a concrete procedure for the optimal adjustment of the gyrocompass follow-up system, utilizing the mathematic model and the stability informations formerly investiaged by the author. For theoptimal determination of the adjustable paramfter K, the performance index(P.I.), ITSE(Intergral of the Time multiplied by the Squared Error) is proposed, namely, P.I. = $\int_{0}^{\infty} t \cdot e^{2}(t)dt$ where t is time and e(t) means control error. Then, the optimal parameter minimizing the performance index is calculated by means of Parseval's theorem and numerical computation, and the validity of the obtained optimal value of the parameter Ka is examined and confirmed through the simulations and experiments. By using, the proposed method, the optimal adjustment can be performed deterministically. But, this can not be expected in the conventional frequency domain analysis. While the Mps of the original system vary to the extent of from 0.98 to 46.27, Mp of the optimal system is evaluated as 1.1 which satisfies the generally accepted frequency domain specification.