• The Korean Journal of Applied Statistics
• /
• v.17 no.1
• /
• pp.27-34
• /
• 2004

The Horvitz-Thompson(H-T) estimator is less efficient than PPS estimators in some cases. We use the two-stage variable transformation in order to remove the drawbacks and increase the efficiency of H-T estimator. We transform the auxiliary variable to use the Midzuno-Sen sampling scheme at the first stage. And the next stage, we also transform the study variable to reduce the variance of H-T estimator using the inclusion probability obtained from the first transformation. We compare the efficiency between a suggested modified H-T estimator and PPS estimators.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.190-190
• /
• 2000

This paper presents a multirate state feedback control (MRSFC) method for systems sensitive to disturbance and noise based on the multirate estimator design using the current estimator. MRSFC updates the controller output slower than the measurement sampling frequency of system output by a lifting factor R=T$\sub$c//T$\sub$s/. The closed-loop MRSFC system is less sensitive to disturbance and noise due to filtering effect than the conventional single-rate control system. The multirate estimator gain is obtained from solving a conventional pole placement problem such that MRSFC has the same spectrum of eigenvalues in the s-plane as the single-rate control. We applied the proposed multirate state feedback controller to a galvanometer servo system. Simulation and experimental results show that settling and tracking performances are improved compared with a conventional single-rate pole placement control (PPC).

• Journal of the Korean Statistical Society
• /
• v.26 no.4
• /
• pp.507-520
• /
• 1997

In this paper we propose a robust Wald-type test which is based on an efficient Mallows-type one-step GM-estimator. The proposed estimator based on the weight function of Song, Park and Nam (1996) has a bounded influence function and a high breakdown point. Under some regularity conditions, we compute the finite-sample breakdown point, and drive asymptotic normality of the proposed estimator. The level and power breakdown points, influence function and asymptotic distribution of the proposed test statistic are main points of this paper. To compare the performance of the proposed test with other tests, we perform some Monte Carlo simulations.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• v.20 no.2
• /
• pp.137-150
• /
• 2016

In this paper we analyze an a posteriori error estimator based on flux recovery for lowest-order finite element discretizations of elliptic interface problems. The flux recovery considered here is based on averaging the discrete normal fluxes and/or tangential derivatives at midpoints of edges with weight factors adapted to discontinuous coefficients. It is shown that the error estimator based on this flux recovery is equivalent to the error estimator of Bernardi and $Verf{\ddot{u}}rth$ based on the standard edge residuals uniformly with respect to jumps of the coefficient between subdomains. Moreover, as a byproduct, we obtain slightly modified weight factors in the edge residual estimator which are expected to produce more accurate results.

• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.89-100
• /
• 1994

In this paper we consider a new estimator of mean residual life (MRL), based on the partial moment of the distribution. The parameters of a partial moment are estimated by its maximum likelihood estimators when the underlying distribution is known. Though the new estimator is not a consistent estimator of the MRL, it is shown to have smaller mean squared error than the well known empirical MRL estimator for certain parametric families. Numerical summaries of the mean squared errors of the new estimator are presented.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.40 no.9
• /
• pp.929-936
• /
• 1991

An estimator for a discrete nonlinear system is derived in the sense of minimum mean square error. An optimal estimator for nonlinear system is very difficult to find and it will be infinite dimensional even if it is found. It has been known that the statistical linearization technique makes it possible to obtain a finite dimensional estimator. In this paper, the procedure of its derivation using the statistical linearization technique that gives an exact mean and variance information is introduced in the sense of minimum mean square error. The derived estimator cannot be clainmed to be globally optimal estimator because it uses the Gaussian assumption to the non-Gaussian distributed nonlinear output. However, the proposed filter exhibits a better performance compared to extended Kalman filter. Simulation results of a simple example present the improvement of the proposed filter in convergent property over the extended Kalman filter.

• Communications for Statistical Applications and Methods
• /
• v.24 no.6
• /
• pp.673-683
• /
• 2017

The least absolute shrinkage and selection operator (LASSO) has been a popular regression estimator with simultaneous variable selection. However, LASSO does not have the oracle property and its robust version is needed in the case of heavy-tailed errors or serious outliers. We propose a robust penalized regression estimator which provide a simultaneous variable selection and estimator. It is based on the rank regression and the non-convex penalty function, the smoothly clipped absolute deviation (SCAD) function which has the oracle property. The proposed method combines the robustness of the rank regression and the oracle property of the SCAD penalty. We develop an efficient algorithm to compute the proposed estimator that includes a SCAD estimate based on the local linear approximation and the tuning parameter of the penalty function. Our estimate can be obtained by the least absolute deviation method. We used an optimal tuning parameter based on the Bayesian information criterion and the cross validation method. Numerical simulation shows that the proposed estimator is robust and effective to analyze contaminated data.

• Journal of the Korean Statistical Society
• /
• v.32 no.1
• /
• pp.85-92
• /
• 2003

We consider the partially linear model E(Y) : X$^{t}$ $\beta$＋η(Z) when the X's are measured with additive error. The semiparametric likelihood estimation ignoring the measurement error gives inconsistent estimator for both $\beta$ and η(.). In this paper we suggest the SIMEX estimator for f to correct the bias induced by measurement error, and explore its properties. We show that the rational linear extrapolant is proper in extrapolation step in the sense that the SIMEX method under this extrapolant gives consistent estimator It is also shown that the SIMEX estimator is asymptotically equivalent to the semiparametric version of the usual parametric correction for attenuation suggested by Liang et al. (1999) A simulation study is given to compare two variance estimating methods for SIMEX estimator.

• Communications for Statistical Applications and Methods
• /
• v.15 no.3
• /
• pp.411-420
• /
• 2008

This paper considers a feasible two-step estimator for seasonal cointegration as the extension of $Br{\ddot{u}}ggeman$ and $L{\ddot{u}}tkepohl$ (2005). It is shown that the reducedrank maximum likelihood(ML) estimator for seasonal cointegration can still produce occasional outliers as that for non-seasonal cointegration even though the sizes of them are not extreme as those in non-seasonal cointegration. The ML estimator(MLE) is compared with the two-step estimator in a small Monte Carlo simulation study and we find that the two-step estimator can be an attractive alternative to the MLE, especially, in a small sample.

• Communications for Statistical Applications and Methods
• /
• v.28 no.4
• /
• pp.315-327
• /
• 2021

The Spatial Autoregressive (SAR) models have drawn considerable attention in recent econometrics literature because of their capability to model the spatial spill overs in a feasible way. While considering the Bayesian analysis of these models, one may face the problem of lack of robustness with respect to underlying prior assumptions. The generalized Bayes estimators provide a viable alternative to incorporate prior belief and are more robust with respect to underlying prior assumptions. The present paper considers the SAR model with a set of linear restrictions binding the regression coefficients and derives restricted generalized Bayes estimator for the coefficients vector. The minimaxity of the restricted generalized Bayes estimator has been established. Using a simulation study, it has been demonstrated that the estimator dominates the restricted least squares as well as restricted Stein rule estimators.