• Journal of the Korean Statistical Society
• /
• v.34 no.2
• /
• pp.141-151
• /
• 2005

Under the assumption that the three-level factors are quantitative, the linear effects are taken more attention than the quadratic effects of the interaction terms. Webb (1971) presented some small incomplete factorial designs that are mixed two- and three-level designs with 20 or fewer runs. The designs provided the estimating linear-by-linear components of interactions between the three-level factors; moreover, they could also offer estimation of interactions that interest the experiments. Webb used ad hoc methods to find these plans; hence, there was still no unified structure to those experiments. In this paper, we develop the methods to construct the $2^{n}3^3$ and $2^{1}3^3$ designs. The designs constructed by these methods not only supply orthogonal estimates of all the main effects but also permit estimation of all the two-factor interactions not involving the quadratic effects. Furthermore, the designs we find are nearly orthogonal.

• Communications for Statistical Applications and Methods
• /
• v.10 no.2
• /
• pp.361-370
• /
• 2003

Those who are interested in making inferences concerning linear combination of valiance components in a simple linear regression model with unbalanced nested error structure can use the confidence intervals proposed in this paper. Two approximate confidence intervals for the sum of two variance components in the model are proposed. Simulation study is peformed to compare the methods. The methods are applied to a numerical example and recommendations are given for choosing a proper interval.

• Communications for Statistical Applications and Methods
• /
• v.27 no.5
• /
• pp.523-533
• /
• 2020

This paper suggests three available methods for finding nonnegative estimates of variance components of the random effects in mixed models. The three proposed methods based on the concepts of projections are called projection method I, II, and III. Each method derives sums of squares uniquely based on its own method of projections. All the sums of squares in quadratic forms are calculated as the squared lengths of projections of an observation vector; therefore, there is discussion on the decomposition of the observation vector into the sum of orthogonal projections for establishing a projection model. The projection model in matrix form is constructed by ascertaining the orthogonal projections defined on vector subspaces. Nonnegative estimates are then obtained by the projection model where all the coefficient matrices of the effects in the model are orthogonal to each other. Each method provides its own system of linear equations in a different way for the estimation of variance components; however, the estimates are given as the same regardless of the methods, whichever is used. Hartley's synthesis is used as a method for finding the coefficients of variance components.

• Proceedings of the KIEE Conference
• /
• 1998.07b
• /
• pp.750-753
• /
• 1998

This paper gives a simple parameterization of all stable unbiased filters to solve the suboptimal mixed $H_2/H_{\infty}$ filtering problem. Using the central filter, mixed $H_2/H_{\infty}$ filter is designed which minimizes the upper bound for the $H_2$ norm of the transfer matrix from a white noise to the estimation error subject to an $H_{\infty}$ norm constraint on the transfer matrix from an energy-bounded noise to the estimation error. The problem of finding suitable estimator gain can be converted into a convex optimization problem involving linear matrix inequalities.

• Asian-Australasian Journal of Animal Sciences
• /
• v.14 no.7
• /
• pp.910-914
• /
• 2001

The teat number of a sow plays an important role for weaning pigs and has been utilized in selection of swine breeding stock. Various linear models have been employed for genetic analyses of teat number although the teat number can be considered as a count trait. Theoretically, Poisson error mixed models are more appropriate for count traits than Normal error mixed models. In this study, the two models were compared by analyzing data simulated with Poisson error. Considering the mean square errors and correlation coefficients between observed and fitted values, the Poisson generalized linear mixed model (PGLMM) fit the data better than the Normal error mixed model. Also these two models were applied to analyzing teat numbers in four breeds of swine (Landrace, Yorkshire, crossbred of Landrace and Yorkshire, crossbred of Landrace, Yorkshire, and Chinese indigenous Min pig) collected in China. However, when analyzed with the field data, the Normal error mixed model, on the contrary, fit better for all the breeds than the PGLMM. The results from both simulated and field data indicate that teat numbers of swine might not have variance equal to mean and thus not have a Poisson distribution.

• Journal of Ecology and Environment
• /
• v.32 no.1
• /
• pp.61-65
• /
• 2009

We compared two nondestructive methods for leaf area estimation using leaves of 16 common plant species classified into six types depending on leaf shape. Relatively good linear relationships between actual leaf area (LA) and leaf length (L), width (W), or the product of length and width (LW) were found for ordinary leaves with lanceolate, oblanceolate, linear and sagitttate shapes with entire margins, serrate margins, mixed margins with a entire form and shallow lobes, and ordinary incised margins. LA was better correlated with LW than L or W, with $R^2$ > 0.91. However, for deeply incised lobes, LA estimation using LW showed low correlation coefficient values, indicating low accuracy. On the other hand, a method using photographic paper showed a good correlation between estimates of area based on the mass of a cut-out leaf image on a photographic sheet (PW) and actual leaf area for all types of leaf shape. Thus, the PW method for LA estimation can be applied to all shapes of leaf with high accuracy. The PW method takes a little more time and has a higher cost than leaf estimation methods using LW based on leaf dimensions. These results indicate that researchers should choose their nondestructive LA estimation method according to their research goals.

• Communications for Statistical Applications and Methods
• /
• v.16 no.6
• /
• pp.971-978
• /
• 2009

Poisson generalized linear mixed models(GLMMs) have been widely used for the analysis of clustered or correlated count data. For the inference marginal likelihood, which is obtained by integrating out random effects is often used. It gives maximum likelihood(ML) estimator, but the integration is usually intractable. In this paper, we propose how to obtain the ML estimator via Laplace approximation based on hierarchical-likelihood (h-likelihood) approach under the Poisson GLMMs. In particular, the h-likelihood avoids the integration itself and gives a statistically efficient procedure for various random-effect models including GLMMs. The proposed method is illustrated using two practical examples and simulation studies.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.575-585
• /
• 2016

Longitudinal studies repeatedly measure outcomes over time. Therefore, repeated measurements are serially correlated from same subject (within-subject variation) and there is also variation between subjects (between-subject variation). The serial correlation and the between-subject variation must be taken into account to make proper inference on covariate effects (Diggle et al., 2002). However, estimation of the covariance matrix is challenging because of many parameters and positive definiteness of the matrix. To overcome these limitations, we propose autoregressive moving average Cholesky decomposition (ARMACD) for the linear mixed models. The ARMACD allows a class of flexible, nonstationary, and heteroscedastic models that exploits the structure allowed by combining the AR and MA modeling of the random effects covariance matrix. We analyze a real dataset to illustrate our proposed methods.

• Communications for Statistical Applications and Methods
• /
• v.27 no.3
• /
• pp.349-363
• /
• 2020

Penalized spline is a suitable nonparametric approach in estimating mean model in small area. However, application of the approach in informative sampling in a published article is uncommon. We propose a semiparametric mixed-model using penalized spline under informative sampling to estimate mean of small area. The response variable is explained in terms of mean model, informative sample effect, area random effect and unit error. We approach the mean model by penalized spline and utilize a penalized spline function of the inclusion probability to account for the informative sample effect. We determine the best and unbiased estimators for coefficient model and derive the restricted maximum likelihood estimators for the variance components. A simulation study shows a decrease in the average absolute bias produced by the proposed model. A decrease in the root mean square error also occurred except in some quadratic cases. The use of linear and quadratic penalized spline to approach the function of the inclusion probability provides no significant difference distribution of root mean square error, except for few smaller samples.

• Journal of Korea Water Resources Association
• /
• v.56 no.8
• /
• pp.509-520
• /
• 2023

Various drought indices are widely used for assessing drought conditions which are affected by many factors such as precipitation, soil moisture, and runoff. The values of drought indices varies depending on hydro-meteorological data and calculation formulas, and the judgment of the drought condition may also vary. This study selected four calculation components such as precipitation data length, accumulation period, probability distribution function, and parameter estimation method as the sources of uncertainty in the calculation of standardized precipitation index (SPI), and evaluated their contributions to the uncertainty using root mean square error (RMSE) and linear mixed model (LMM). The RMSE estimated the overall errors in the SPI calculation, and the LMM was used to quantify the uncertainty contribution of each factor. The results showed that as the accumulation period increased and the data period extended, the RMSEs decreased. The comparison of relative uncertainty using LMM indicated that the sample size had the greatest impact on the SPI calculation. In addition, as sample size increased, the relative uncertainty related to the sample size used for SPI calculation decreased and the relative uncertainty associated with accumulation period and parameter estimation increased. In conclusion, to reduce the uncertainty in the SPI calculation, it is essential to collect long-term data first, followed by the appropriate selection of probability distribution models and parameter estimation methods that represent well the data characteristics.