• Communications for Statistical Applications and Methods
• /
• 제9권3호
• /
• pp.595-605
• /
• 2002

The central composite design is widely used for estimating second order response surfaces. This type of design is composed of $2^k$ factorial points, axial points and center points. In this paper, we suggest a version of central composite design where the positions of the axial points are indicated by two numbers, and study properties of this design. We obtain the variances and covariances of the estimators of the regression coefficients. Conditions are obtained for this design to be orthogonal and rotatable. This design is compared with other designs on the basis of efficiency.

• Journal of the Korean Statistical Society
• /
• 제26권1호
• /
• pp.1-22
• /
• 1997

In the linear regression model $y_{i}$ = .alpha. $x_{i}$ $^{T}$ .beta. + .epsilon.$_{i}$ , i = 1,2,...,n, the weighted pairwise absolute deviation (WPAD) estimator was defined by minimizing the dispersion function D (.beta.) = .sum..sum.$_{{i $w_{{ij}}$$\mid$ $r_{j}$ (.beta.) $r_{i}$ (.beta.)$\mid$, where $r_{i}$ (.beta.)'s are residuals and $w_{{ij}}$'s are weights. This estimator can achive bounded total influence with positive breakdown by choice of weights $w_{{ij}}$. In this paper, we consider a more general type of dispersion function than that of D(.beta.) and propose a pairwise GM-estimator based on the dispersion function. Under some regularity conditions, the proposed estimator has a bounded influence function, a high breakdown point, and asymptotically a normal distribution. Results of a small-sample Monte Carlo study are also presented. presented.

• Communications for Statistical Applications and Methods
• /
• 제8권2호
• /
• pp.357-366
• /
• 2001

The present paper develops a method to check the propriety of link functions for binary data. In order to parameterize a certain type of goodness of the link, a family of link functions indexed by a shape parameter is proposed. I first investigate the maximum likelihood estimation of the shape parameter as well as regression parameters and then derive their large sample behaviors of the estimators. A score test is considered to evaluate the goodness of the current link function. For illustration, I employ two families of power transformations, the modulus transformation by John and Draper (1980) and the extended power transformation by Yeo and Johnson (2000), which are appropriate to detect symmetric and asymmetric inadequacy of the selected link function. respectively.

• 한국데이터정보과학회:학술대회논문집
• /
• 한국데이터정보과학회 2003년도 춘계학술대회
• /
• pp.1-14
• /
• 2003

In man cases, the measurement error variances may be functions of the unknown true values or related covariate. In some cases, the measurement error variances increase in proportion to the value of predictor. This paper develops estimators of the parameters of a linear measurement error variance function under stratified multistage random sampling design and additional conditions. Also, this paper evaluates and compares the power of an asymptotically unbiased test with that of an asymptotically biased test. The proposed method are applied to blood sample measurements from the U.S. Third National Health and Nutrition Examination Survey(NHANES III)

• 천문학회보
• /
• 제37권1호
• /
• pp.35.1-35.1
• /
• 2012

Determining the virial factor of the broad-line region (BLR) gas is crucial in calibrating AGN black hole mass estimators, since the measured line-of-sight velocity needs to be converted into the representative velocity of the BLR gas. The unknown virial factor has been empirically calibrated based on the $M_{BH}-{\sigma}_*$ relation of non-AGN galaxies, but the claimed values are different by a factor of 2 in recent studies. We investigate the origin of the difference by measuring the $M_{BH}-{\sigma}_*$ relation using the most updated nearby galaxy sample, and explore the dependence of the virial factor on the various fitting methods. We find that the discrepancy is mostly caused by the sample bias while the difference stemming from various regression methods is marginal. Based on the best-determined virial factor, we present the updated $M_{BH}-{\sigma}_*$ relation of local active galaxies.

• 품질경영학회지
• /
• 제23권3호
• /
• pp.140-155
• /
• 1995

This paper proposes a quality control method of assembly parts based on parts inspection information. We use the parts inspection information for reduction of defective fraction in assembly and prediction of assembly quality. At first, we build the functional relationship between parts and assembled unit by regression analysis. Secondly, if we use the inspection informations of parts which are estimators of means and variances, we can predict the nonconforming probability of assembled unit and propose the best assembly method. In a case study, we showed the efficency of the proposed method in two part assembly.

• 응용통계연구
• /
• 제17권1호
• /
• pp.145-151
• /
• 2004

SAS의 PROC MIXED procedure는 다양한 형태의 혼합모형에 적합한 자료를 분석하고, 그 자료들이 채집된 모집단의 모수들에 관한 통계적 추론을 하는데 사용된다. 그러나 혼합모형에 해당되는 불균형중첩오차구조를 갖는 선형회귀모형안에 나타나는 두개의 분산의 합에 대한 신뢰구간을 구할 때 PROC MIXED의 REML추정량으로부터 계산되는 신뢰구간은 신뢰계수를 지키지 못한다는 것을 시뮬레이션을 통하여 보인다.

• Journal of the Korean Data and Information Science Society
• /
• 제16권4호
• /
• pp.1017-1026
• /
• 2005

We consider the problem of modeling count data where the observation period is determined by the survival time of the individual under study. We assume random effects or frailty model to allow for a possible association between the death times and the counts. We assume that, given a random effect, the death times follow a Weibull distribution with a rate that depends on some covariates. For the counts, given the random effect, a Poisson process is assumed with the intensity depending on time and the covariates. A gamma model is assumed for the random effect. Maximum likelihood estimators of the model parameters are obtained. The model is applied to data set of patients with breast cancer who received a bone marrow transplant. A model for the time to death and the number of supportive transfusions a patient received is constructed and consequences of the model are examined.

• Communications for Statistical Applications and Methods
• /
• 제19권6호
• /
• pp.809-817
• /
• 2012

We propose a semiparametric inference for a generalized varying coefficient partially linear model(VCPLM) for negative binomial data. The VCPLM is useful to model real data in that varying coefficients are a special type of interaction between explanatory variables and partially linear models fit both parametric and nonparametric terms. The negative binomial distribution often arise in modelling count data which usually are overdispersed. The varying coefficient function estimators and regression parameters in generalized VCPLM are obtained by formulating a penalized likelihood through smoothing splines for negative binomial data when the shape parameter is known. The performance of the proposed method is then evaluated by simulations.

• Journal of the Korean Statistical Society
• /
• 제24권2호
• /
• pp.303-312
• /
• 1995

We consider the following type of general semi-parametric non-linear regression model : $y_i = f_i(\theta) + \epsilon_i, i=1, \cdots, n$ where ${f_i(\cdot)}$ represents the set of non-linear functions of the unknown parameter vector $\theta' = (\theta_1, \cdots, \theta_p)$ and ${\epsilon_i}$ represents the set of measurement errors with unknown distribution. Under suitable finite-sample, small-dispersion asymptotic framework, we derive a general lower bound for the asymptotic mean squared error (AMSE) matrix of the Gauss-consistent estimator of $\theta$. We then prove the fundamental result that the general non-linear least squares estimator (NLSE) is an optimal estimator within the class of all regular Gauss-consistent estimators irrespective of the type of the distribution of the measurement errors.