• 한국데이터정보과학회:학술대회논문집
• /
• 2003.10a
• /
• pp.93-100
• /
• 2003

It is necessary to check the curvature of selected covariates in regression diagnostics. There are various graphical methods using residual plots based on least squares fitting. The sensitivity of LS fitting to outliers can distort their residuals, making the identification of the unknown function difficult to impossible. In this paper, we compare combining conditional expectation and residual plots(CERES Plots) between least square fit and robust fits using Huber M-estimator. Robust CERES will be far less distorted than their LS counterparts in the presence of outliers and hence, will be more useful in identifying the unknown function.

• The Korean Journal of Applied Statistics
• /
• v.25 no.3
• /
• pp.403-413
• /
• 2012

The modifications suggested in Uhm et al. (2011) are studied using a partly parametric version of Aalen's additive risk model. A follow-up time period is partitioned into intervals, and hazard functions are estimated as a piecewise constant in each interval. A maximum likelihood estimator by iteratively reweighted least squares and variance estimates are suggested based on the model as well as evaluated by simulations using mean square error and a coverage probability, respectively. In conclusion the modifications are needed when there are a small number of uncensored deaths in an interval to estimate the piecewise constant hazard function.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.5
• /
• pp.1153-1165
• /
• 2017

Structural equation modeling (SEM) is a basic tool for testing theories in a variety of disciplines. A maximum likelihood (ML) method for parameter estimation is by far the most widely used in SEM. Alternatively, two-stage least squares (2SLS) estimator has been proposed as a more robust procedure to address model misspecification. A regularized extension of 2SLS, two-stage ridge least squares (2SRLS) has recently been introduced as an alternative to ML to effectively handle the small-sample-size issue. However, it is unclear whether and when misspecification and small sample sizes may pose problems in theory testing with 2SLS, 2SRLS, and ML. The purpose of this article is to evaluate the three estimation methods in terms of inferences errors as well as parameter recovery under two experimental conditions. We find that: 1) when the model is misspecified, 2SRLS tends to recover parameters better than the other two estimation methods; 2) Regardless of specification errors, 2SRLS produces small or relatively acceptable Type II error rates for the small sample sizes.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.5 no.3
• /
• pp.169-173
• /
• 2012

This paper investigates the performance of Turbo equalization in wireless multipath channels. Turbo equalization mainly consists of a SISO(soft-in soft-out) equalizer and a SISO decoder. Iterative channel estimators can improve the accuracy of channel estimates by soft information fed back from the SISO decoder. Comparing iterative channel estimators with LMS(least mean square) and RLS(recursive least squares) algorithms, which are the most common algorithms to estimate and track a time-varying channel impulse response, the iterative channel estimator with RLS converges more faster than the one with LMS. However, the difference of BER(bit error rate) performances gradually decreases as the number of iterations for Turbo equalization increases.

• Communications for Statistical Applications and Methods
• /
• v.24 no.5
• /
• pp.519-531
• /
• 2017

We consider goodness-of-fit test statistics for Weibull distributions when data are randomly censored and the parameters are unknown. Koziol and Green (Biometrika, 63, 465-474, 1976) proposed the $Cram\acute{e}r$-von Mises statistic's randomly censored version for a simple hypothesis based on the Kaplan-Meier product limit of the distribution function. We apply their idea to the other statistics based on the empirical distribution function such as the Kolmogorov-Smirnov and Liao and Shimokawa (Journal of Statistical Computation and Simulation, 64, 23-48, 1999) statistics. The latter is a hybrid of the Kolmogorov-Smirnov, $Cram\acute{e}r$-von Mises, and Anderson-Darling statistics. These statistics as well as the Koziol-Green statistic are considered as test statistics for randomly censored Weibull distributions with estimated parameters. The null distributions depend on the estimation method since the test statistics are not distribution free when the parameters are estimated. Maximum likelihood estimation and the graphical plotting method with the least squares are considered for parameter estimation. A simulation study enables the Liao-Shimokawa statistic to show a relatively high power in many alternatives; however, the null distribution heavily depends on the parameter estimation. Meanwhile, the Koziol-Green statistic provides moderate power and the null distribution does not significantly change upon the parameter estimation.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.23 no.4
• /
• pp.366-370
• /
• 2013

Regression Analysis is an analyzing method of regression model to explain the statistical relationship between explanatory variable and response variables. This paper introduce Theil's method to find a fuzzy regression model which explain the relationship between explanatory variable and response variables. Theil's method is a robust method which is not sensive to outliers. Theil's method use medians of rate of increment based on randomly chosen pairs of each components of ${\alpha}$-level sets of fuzzy data in order to estimate the coefficients of fuzzy regression model. We propose an example to show Theil's estimator is robust than the Least squares estimator.

• The Journal of Asian Finance, Economics and Business
• /
• v.7 no.11
• /
• pp.529-540
• /
• 2020

The objective of this research paper is to study the simultaneous relationship between fiscal decentralization, corruption, and income inequality among Vietnamese provinces. We use a balanced panel data set of 63 provinces/cities in Vietnam in the period from 2011 to 2018. The study used 3SLS-GMM (Three Stage Least Squares - Generalized Method of Moments estimator) and GMM-HAC (Generalized Method of Moments - Heteroskedastic and Autocorrelation Consistent estimator). Empirical evidence shows a strong simultaneous relationship: increased corruption will increase regional income disparities, income inequality, and increase fiscal decentralization. In addition, the results also suggest that an increase in per-capita income will reduce the level of corruption, or better control corruption of each province. The degree of increase in income inequality, which reduces fiscal decentralization, is the same for trade liberalization. All demonstrate that there is a simultaneous relationship between fiscal decentralization, corruption, and income inequality. In a region of high public governance quality, fiscal decentralization positively effects its economic growth. This issue will indirectly increase income inequality between provinces within a country. Our findings imply that a country's fiscal decentralization strategy should be linked to improving corruption control and local governance effectiveness, indirectly improving income inequality between localities or regions.

• Journal of the Korean Statistical Society
• /
• v.22 no.2
• /
• pp.271-282
• /
• 1993

Unit root test for temporally aggregated first order autoregressive process is considered. The temporal aggregate of fist order autoregression is an autoregressive moving average of order (1,1) with moving average parameter being function of the autoregressive parameter. One-step Gauss-Newton estimators are proposed and are shown to have the same limiting distribution as the ordinary least squares estimator for unit root when complete observations are available. A Monte-Carlo simulation shows that the temporal aggregation have no effect on the size. The power of the suggested test are nearly the same as the powers of the test based on complete observations.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1553-1556
• /
• 1997

In the paper, we porpose a new mehtod of extending PLS(Partial Least Squares) regressiion method to nonlinear framework and apply it to the estimation of product compositions in high-purity distillation column. There have veen similar efforets to overcome drawbacks of PLS by using nonlinear-mapping ability of meural networks, however, they failed to show great improvement over PLS since they focused only in capturing nonlinear functional relationship between input data, not on nonlinear correlation inthe data set. By incorporating the structure of Robust Auto Associative Networks(RAAN) into that of previous nonlinear PLS, we can handle nonlinear correlation as well as nonlinear functional relationship. The application result shows that the proposed method performs better than previous ones even for nonlinearities caused by changing operating conditions, limited observations, and existence of meas-unrement noises.

• 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.