• The Korean Journal of Applied Statistics
• /
• v.21 no.2
• /
• pp.327-340
• /
• 2008

This paper derives joint and conditional Lagrange multiplier tests based on information matrix for testing functional form and/or the presence of autocorrelation in a regression model. Small sample properties of these tests are assessed by Monte Carlo study and comparisons are made with LM tests based on Hessian matrix. The results show that the proposed $LM_E$ tests have the most appropriate finite sample performance.

• Journal for History of Mathematics
• /
• v.11 no.2
• /
• pp.63-76
• /
• 1998

When teaching the linear regression analysis in the class, the power transformation must be introduced to fit the linear regression model for nonlinear data. Box and Cox (1964) proposed the attractive power transformation technique which is so called Box-Cox transformation. In this paper, an effective algorithm selecting an appropriate value for Box-Cox transformation is developed which is considered to find a value minimizing error sum of squares. When the proposed algorithm is used to find a value for transformation, the number of iterations needs to be considered. Thus, the number of iterations is examined through simulation study. Since SAS is one of most widely used packages and does not provide the procedure that performs iterative Box-Cox transformation, a SAS program automatically choosing the value for transformation is developed. Hence, the students could learn how the Box-Cox transformation works, moreover, researchers can use this for analysis of data.

• Journal of the Korean Data and Information Science Society
• /
• v.12 no.1
• /
• pp.103-125
• /
• 2001

The Box-Cox transformation is a well known family of power transformation that brings a set of data into agreement with the normality assumption of the residuals and hence the response variable of a postulated model in regression analysis. This paper proposes a new method for estimating the Box-Cox transformation using maximization of the Sample Entropy statistic which forces the data to get closer to normal as much as possible. A comparative study of the proposed procedure with the maximum likelihood procedure, the procedure via artificial regression estimation, and the recently introduced maximization of the Shapiro-Francia W' statistic procedure is given. In addition, we generate a table for the optimal spacings parameter in computing the Sample Entropy statistic.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.5
• /
• pp.967-976
• /
• 2011

The Box-Cox transformation is a well known family of power transformations that brings a set of data into agreement with the normality assumption of the residuals and hence the response variable of a postulated model in regression analysis. Normalization (studentization) of the regressors is a common practice in analyzing microarray data. Here, we implement Box-Cox transformation in normalizing regressors in microarray data. Pridictabilty of the model can be improved using data transformation compared to studentization.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.29 no.1
• /
• pp.143-150
• /
• 2006

Box-Cox model and T-factor method have been widely used to measure economic depreciations for industrial property. The Box-Cox model which combines economic efficiency with depreciation pattern is here extended to the reliability function. To do so a Rayleigh distribution which has been used to estimate the reliability of current assets was chosen as an efficiency curve of marginal productivity. Such an approach provides the possibility to classify the efficiency curves into four categories. It is also possible to analyze the types of depreciation curves. Therefore, the power family of a non-linear Box-Cox model could be set at certain constant values, then the model can be transformed into a linear model to estimate the economic depreciation rates by utilizing the reliability function. Estimating the resultant linear regression equation requires minimal number of observations, while at the same time facilitating the test of hypothesis on depreciation rates.

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.351-364
• /
• 2003

This paper derives Lagrange multiplier tests based on Double-Length Artificial Regression and Outer-Product Gradient for testing linear and log-linear panel regressions against Box-Cox alternatives. The proposed DLR based LM tests are easy to implement in an error component model. From the Monte Carlo study, the DLR based LM tests are recommended for testing functiona forms.

• The Korean Journal of Applied Statistics
• /
• v.19 no.3
• /
• pp.461-479
• /
• 2006

This paper derives joint and conditional Lagrange multiplier tests based on Double-Length Artificial Regression(DLR) for testing functional form and/or the presence of individual(time) effect in a panel regression model. Small sample properties of these tests are assessed by Monte Carlo study, and comparisons are made with LM tests based on Outer Product Gradient(OPG). The results show that the proposed DLR based LM tests have the most appropriate finite sample performance.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1191-1204
• /
• 2017

Weighted least squares (WLS) estimation is often easily used for the data with heteroscedastic errors because it is intuitive and computationally inexpensive. However, WLS estimator is less robust to a few outliers and sometimes it may be inefficient. In order to overcome robustness problems, Box-Cox transformation, Huber's M estimation, bisquare estimation, and Yohai's MM estimation have been proposed. Also, more efficient estimations than WLS have been suggested such as Bayesian methods (Cepeda and Achcar, 2009) and semiparametric methods (Kim and Ma, 2012) in heteroscedastic error models. Recently, Çelik (2015) proposed the weight methods applicable to the heteroscedasticity patterns including butterfly-distributed residuals and megaphone-shaped residuals. In this paper, we review heteroscedastic regression estimators related to robust or efficient estimation and describe their properties. Also, we analyze cost data of U.S. Electricity Producers in 1955 using the methods discussed in the paper.

• Journal of the Korean GEO-environmental Society
• /
• v.10 no.7
• /
• pp.7-14
• /
• 2009

The correlation equation empirically proposed to obtain compression indexes has been proposed to conveniently obtain the value using the soil parameter that can be obtained through simple tests when the number of time of consolidation testing is low or the distribution is large but most of the analyzed regions are limited to certain regions abroad or in the country and multiple data were integrated for use in many cases, thus it is not very reasonable to apply it. Therefore, to establish a new design method considering the uncertainty of the ground, it was selected the Kwangyang port area of which the data have been collected recently thus are relatively more reliable as the subject region of the study in order to maximally reduce the uncertainty of test data. After performing the verification of the normality of the consolidation test data obtained from the selected region and the transformation of variables, a prediction formula was proposed through the regression model with the transformed variables and the proposed regression model with transformed variables was compared with existing empirical equations to verify the suitability of the proposed model formula. After analyzing, it was confirmed that the coefficient of determination was increased after the Box-Cox variable transformation, thus the explanatory power was being enhanced and through the root-mean-square-error method, it was confirmed that the proposed model formula showed the most closed value to the test value.

• The Korean Journal of Applied Statistics
• /
• v.25 no.1
• /
• pp.205-213
• /
• 2012

Transforming response variable is a general tool to adapt data to a linear regression model. However, it is well known that response transformations in linear regression are very sensitive to one or a few outliers. Many methods have been suggested to develop transformations that will not be influenced by potential outliers. Recently Cheng (2005) suggested to using a trimmed likelihood estimator based on the idea of the least trimmed squares estimator(LTS). However, the method requires presetting the number of outliers and needs many computations. A new method is proposed, that can solve the problems addressed and improve the robustness of the estimates. The method uses a stepwise procedure, suggested by Hadi and Simonoff (1993), to detect outliers that determine response transformations.