• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1990년도 하계학술대회 논문집
• /
• pp.72-76
• /
• 1990

A new statistical linearization technique for nonlinear system called covariance matching method is proposed in this paper. The covariance matching method makes the mean and variance of an approximated output be identical real functional output, and the distribution of the approximated output have identical shape with a given random input. Also, the covariance matching method can be easily implemented for statistical analysis of nonlinear systems with a combination of linear system covariance analysis.

• 응용통계연구
• /
• 제27권5호
• /
• pp.755-771
• /
• 2014

등방성(isotropy)은 공분산 모형(covariance model)에 기반으로 공간 예측(spatial prediction)이라 불리우는 크리깅(kriging) 을 용이하게 수행하기 위한 주요 가정 중의 하나로 알려져있다. 공간 과정에서 등방성이 충족되지 않는 경우에는, 보다 신뢰성 예측을 생성하기 위해 비등방성 공분산 모형(covariance model)과 관련된 모수들(각도 및 비율)를 추정해야 한다. 본 논문에서는 여러 방향의 기하학적 비등방성 모형(geometrically anisotropic covariance models)의 가중 평균으로 표현되는 확장된 형태의 기하학적 비등방성(geometrically extended anisotropic) 공분산모형을 제안한다. 연구에 관심이 되는 모수를 추정하기 위해 최대우도추정법(maximum likelihood estimation method)을 이용하였다. 제안한 모형의 성능을 평가하기 위해 등방성 공분산모형과 기하학적 비등방성 모형을 고려한 모의실험을 수행하였다. 또한 확장된 기하학적 비등방성 모형을 적용한 미세먼지 농도자료 분석을 실시하였다.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2002년도 ITC-CSCC -1
• /
• pp.18-21
• /
• 2002

We present a phase covariance model that can well represent stimulus intensity as well af feature binding (i.e., covariance). The model is represented by complex neural equations, which is a mean field model of stochastic neural model such as Boltzman machine and sigmoid belief networks.

• 한국수학교육학회지시리즈B:순수및응용수학
• /
• 제13권2호
• /
• pp.137-149
• /
• 2006

We introduce the notion of the generalized covariance and variance for bounded linear operators on Hilbert space, and prove that the generalized covariance-variance inequality holds. It turns out that the inequality is a useful formula in tile study of inequality involving linear operators in Hilbert spaces.

• Journal of applied mathematics & informatics
• /
• 제28권3_4호
• /
• pp.913-921
• /
• 2010

The derivative influence measure is adapted to the Cholesky decomposition of a covariance matrix. Formulas for the derivative influence of observations on the Cholesky root and the inverse Cholesky root of a sample covariance matrix are derived. It is easy to implement this influence diagnostic method for practical use. A numerical example is given for illustration.

• 대기
• /
• 제23권1호
• /
• pp.33-37
• /
• 2013

In meteorology, exploitation of Kalman filter as a data assimilation system is virtually impossible due to simultaneous requirements of adjoint model and large computer resource. The other substitute of utilizing ensemble Kalman filter is only affordable by compensating an enormous usage of computing resource. Furthermore, the latter employs ensemble integration sets for evolving the background error covariance matrix by compensating the dynamical feature of the temporal evolution of weather conditions. We propose a new implementation method that works without the adjoint model by utilizing the explicit expression of the background error covariance matrix in backward evolution. It will also break a barrier in the evolution of the covariance matrix. The method may be applied with a slight modification to the real time assimilation or the retrospective analysis.

• 제어로봇시스템학회논문지
• /
• 제15권1호
• /
• pp.31-36
• /
• 2009

This paper presents a novel scalar adaptive filter, which is reformulated by additional acceleration term. The filter continuously estimates three different kinds of covariance such as the measurement noise covariance, the velocity error covariance and the acceleration error covariance. For estimating three covariances, we use the innovation method for the measurement noise covariance and the least square method for other covariances. In order to verify the proposed filter performance compared with the conventional scalar adaptive filter, we make indoor experimental environment similar to outdoor test using the ultrasonic sensors instead of GPS. Experimental results show that the proposed filter has better position accuracy than the traditional scalar adaptive filter.

• Journal of Preventive Medicine and Public Health
• /
• 제6권1호
• /
• pp.77-79
• /
• 1973

The Mechanism for sorting out the covariance effect is known as the covariance analysis. The sorting out of regression and correlation effect is an obvious application of the covariance analysis. The result of Do population by age groups (15 ages interval) from 1966 Census and from 1970 Census has been applied to analyzing covariability by the analysis of covariance. The results are as follows. (1) The signicance of the regression of 1970 population on 1966 population is assured as F=116.5 (2) There is a significant difference between mean of each age group. (F=88.1) (3) There is very little evidence of significant heterogeneity of regression between age groups. (F=0.72)

• 품질경영학회지
• /
• 제29권1호
• /
• pp.1-10
• /
• 2001

We consider the problem of detecting special variations in multivariate $T^2$-control chart when two or more multivariate outliers are present. Since a multivariate outlier may reflect slippage in mean, variance, or correlation, it can distort the sample mean vector and sample covariance matrix. Damaged sample mean vector and sample covariance matrix have difficulty in examining special variations clearly, An alternative to detection outliers or special variations is to use robust estimators of mean vector and covariance matrix that are less sensitive to extreme observations than are the standard estimators $\bar{x}$ and $\textbf{S}$. We applied popular minimum volume ellipsoid(MVE) and minimum covariance determinant(MCD) method to estimate mean vector and covariance matrix and compared its results with standard $T^2$-control chart using simulated multivariate data with outliers. We found that the modified $T^2$-control chart based on the above robust methods were more effective in detecting special variations clearly than the standard $T^2$-control chart.

• Journal of the Korean Data and Information Science Society
• /
• 제23권4호
• /
• pp.807-814
• /
• 2012

We know that the exponentially weighted moving average (EWMA) control charts are sensitive to detecting relatively small shifts. Multivariate EWMA control charts are considered for monitoring of variance-covariance matrix when the distribution of process variables is multivariate normal. The performances of the proposed EWMA control charts are evaluated in term of average run length (ARL). The performance is investigated in three types of shifts in the variance-covariance matrix, that is, the variances, covariances, and variances and covariances are changed respectively. Numerical results show that all multivariate EWMA control charts considered in this paper are effective in detecting several kinds of shifts in the variance-covariance matrix.