• Communications for Statistical Applications and Methods
• /
• v.3 no.3
• /
• pp.39-46
• /
• 1996

Most of existing nonparametric test statistics are based on the residuals which are obtained by regressing the data to a parametric model. In this paper we compare power of goodness-of-fit test statistics for testing the (null)parametric model versus the (alternative) nonparametric model.

• Communications for Statistical Applications and Methods
• /
• v.4 no.1
• /
• pp.259-269
• /
• 1997

Bayesian inference for a record value statistics(RVS) model of nonhomogeneous Poisson process is considered. We seal with Bayesian inference for double exponential, Gamma, Rayleigh, Gumble RVS models using Gibbs sampling and Metropolis algorithm and also explore Bayesian computation and model selection.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.33-36
• /
• 1999

This paper proposes a mathematical model for the statistics of the W-CDMA signals with different bandwidth. Based on the statistics of numerically generated signals, a mathematical model is obtained such as Rayleigh, Rician, and Maxwell distribution. We employ Chi-square test to verify the fitness of the mathematical model with signal statistics. The results show obviously that the new proposed model is useful for representing W-CDMA signals.

• The Korean Journal of Applied Statistics
• /
• v.28 no.2
• /
• pp.361-369
• /
• 2015

A general linear mixed-effects model is often used to analyze repeated measurement experiment data of a continuous response variable. However, a general linear mixed-effects model can give improper analysis results when simultaneously detecting heteroscedasticity and the non-normality of population distribution. To achieve a more robust estimation, we used a heavy-tailed linear mixed-effects model for a more exact and reliable analysis conclusion than a general linear mixed-effects model. We also provide reliability analysis results for further research.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.2
• /
• pp.543-557
• /
• 2006

This paper is concerned with the applicability of the chi-square approximation to the six disparity statistics: the Pearson chi-square, the generalized likelihood ratio, the power divergence, the blended weight chi-square, the blended weight Hellinger distance, and the negative exponential disparity statistic. Three dimensional contingency tables of small and moderate sample sizes are generated to be fitted to all possible hierarchical log-linear models: the completely independent model, the conditionally independent model, the partial association models, and the model with one variable independent of the other two. For models with direct solutions of expected cell counts, point estimates and confidence intervals of the 90 and 95 percentage points of six statistics are explored. For model without direct solutions, the empirical significant levels and the empirical powers of six statistics to test the significance of the three factor interaction are computed and compared.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.859-871
• /
• 2003

In this paper we introduce some adaptive M-estimators using selector statistics to estimate the slope of regression model under the symmetric and continuous underlying error distributions. This selector statistics is based on the residuals after the preliminary fit L$_1$ (least absolute estimator) and the idea of Hogg(1983) and Hogg et. al. (1988) who used averages of some order statistics to discriminate underlying symmetric distributions in the location model. If we use L$_1$ as a preliminary fit to get residuals, we find the asymptotic distribution of sample quantiles of residual are slightly different from that of sample quantiles in the location model. If we use the functions of sample quantiles of residuals as selector statistics, we find the suitable quantile points of residual based on maximizing the asymptotic distance index to discriminate distributions under consideration. In Monte Carlo study, this adaptive M-estimation method using selector statistics works pretty good in wide range of underlying error distributions.

• Journal of the Korean Data and Information Science Society
• /
• v.6 no.2
• /
• pp.29-39
• /
• 1995

A model used in the past may be altered at will in modeling for the future. For this situation, the multiprocess dynamic model provides a general framework. In this paper we consider the multiprocess discount normal model with parameters having a time dependent non-linear structure. This model has nice properties such as insensitivity to outliers and quick reaction to abrupt changes of pattern.

• Communications for Statistical Applications and Methods
• /
• v.19 no.6
• /
• pp.859-868
• /
• 2012

Due to a rapidly aging society, the future Korea mortality rate is important for planning national financial strategies and social security policies. Old age mortality statistics are very limited in their ability to project a future mortality rate; therefore, it is essential to accurately estimate the old age mortality rate. In this paper, we show that the CK model with a Relational model as a base model provides accurate estimates of old age mortality rates.

• The Korean Journal of Applied Statistics
• /
• v.25 no.3
• /
• pp.457-464
• /
• 2012

Small area estimation is a statistical inference method to overcome the large variance due to the small sample size allocated in a small area. Recently some nonparametric estimators have been applied to small area estimation. In this study, we suggest a nonparametric mixed effect small area estimator using kernel smoothing and compare the small area estimators using labor statistics.

• Communications for Statistical Applications and Methods
• /
• v.3 no.3
• /
• pp.31-37
• /
• 1996

A hierarchical loglinear model comparison method is developed which is based on the well kmown partitioned likelihood ratio statistiss. For any paels, we can regard the difference of the geedness of fit statistics as the variation explained by a full model, and develop a partial test to compare a full model with a reduced model in that hierarchy. Note that this has similar arguments as that of the regression analysis.