• Communications for Statistical Applications and Methods
• /
• v.15 no.1
• /
• pp.95-104
• /
• 2008

This paper addresses issues of robustness in Bayesian optimal design. We may have difficulty applying Bayesian optimal design principles because of the uncertainty of prior distribution. When there are several plausible prior distributions and the efficiency of a design depends on the unknown prior distribution, robustness with respect to misspecification of prior distribution is required. We suggest a new optimal design criterion which has relatively high efficiencies across the class of plausible prior distributions. The criterion is applied to the problem of estimating the turning point of a quadratic regression, and both analytic and numerical results are shown to demonstrate its robustness.

• Communications for Statistical Applications and Methods
• /
• v.19 no.1
• /
• pp.169-175
• /
• 2012

The sensitivity of the performance measures such as the mean and the standard deviation of the queue length and the blocking probability with respect to the moments of the service time are numerically investigated. The service time distribution is fitted with phase type(PH) distribution by matching the first three moments of service time and the M/G/c retrial queue is approximated by the M/PH/c retrial queue. Approximations are compared with the simulation results.

• Communications for Statistical Applications and Methods
• /
• v.23 no.5
• /
• pp.411-421
• /
• 2016

Large frequency limiting distributions of two errors in realized covariance are investigated under noisy and non-synchronous high frequency sampling situations. The first distribution characterizes increased variance of the realized covariance due to noise for large frequency and the second distribution characterizes decreased variance of the realized covariance due to discretization for large frequency. The distribution of the combined error enables us to determine the sampling frequency which depends on a nuisance parameter. A consistent estimator of the nuisance parameter is proposed.

• The Korean Journal of Applied Statistics
• /
• v.22 no.5
• /
• pp.1097-1102
• /
• 2009

Classification of subjects with unknown distribution in small sample size setup may involve order-restricted constraints in multivariate parameter setups. Those problems makes optimality of conventional likelihood ratio based statistical inferences not feasible. Fortunately, Roy (1953) introduced union-intersection principle(UIP) which provides an alternative avenue. Redescending M-estimator along with that principle yields a considerably appropriate robust testing procedure. Furthermore, conditionally distribution-free test based upon exact permutation theory is used to generate p-values, even in small sample. Applications of this method are illustrated in simulated data and read data example (Lobenhofer et al., 2002)

• Communications for Statistical Applications and Methods
• /
• v.6 no.1
• /
• pp.151-158
• /
• 1999

Distribution-free tests are proposed for AB/BA 2*2 cross-over design data based on placements introduced by Orban and Wolfe(1992). In this paper we suggest the homogeneity test or carry-over effects and also suggest the test for direct effects and the test for period effects under the same carry-over effects. The properties such as iterative asymptotic distribution for the proposed tests are discussed.

• Communications for Statistical Applications and Methods
• /
• v.6 no.2
• /
• pp.595-600
• /
• 1999

We present a simple geometric approach which uses polar transformation and elementary trigonometry to evaluating an orthant probability in a bivariate normal distribution. Figures are provided to illustrate the situation for varying correlation coefficient. We derive the distribution of the sample correlation coefficient from a bivariate normal distribution when the sample size is 2 as an application.

• Communications for Statistical Applications and Methods
• /
• v.11 no.2
• /
• pp.323-327
• /
• 2004

In this paper, we investigate the limiting distribution of the Ljung-Box test statistic in the unit root AR models with ARCH errors. We show that the limiting distribution is approximately chi-square distribution with the degrees of freedom only depending on the number of autocorrelation lags appearing in the test. Some simulation results are provided for illustration.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.909-917
• /
• 2003

Although many classification models have been used to classify binary data, none of the classification models dominates all varying circumstances depending on the number of variables and the size of data(Asparoukhov and Krzanowski (2001)). This paper proposes a classification model which uses information on marginal distributions of sub-variables and its maximum entropy distribution. Classification experiments by using simulation are discussed.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.1007-1015
• /
• 2003

In this paper we introduced a new score generating function for the rank dispersion function in a general linear model. Based on the new score function, we derived the null asymptotic theory of the rank-based hypothesis testing in a linear model. In essence we showed that several rank test statistics, which are primarily focused on our new score generating function and new dispersion function, are mainly distribution free and asymptotically converges to a chi-square distribution.

• Communications for Statistical Applications and Methods
• /
• v.5 no.3
• /
• pp.837-845
• /
• 1998

We shall propose several estimators and confidence intervals for the scale parameter in a uniform distribution with the presence of a unidentified outlier and obtain biases and mean squared errors for their proposed estimators. And we shall numerically compare the performances for the proposed several estimators of the sclae parameter. Also, we shall compare lengths of confidence intervals of the scale parameter in a uniform distribution through Monte Carlo methods.