• The Korean Journal of Applied Statistics
• /
• v.5 no.2
• /
• pp.255-268
• /
• 1992

We propose an iterated interpolation approach for the estimation fo time series parameters in the presence of outliers. The proposed approach iterates the parameter estimation stage and the outlier detection stage until no further outliers are detected. For the detection of outliers, interpolation diagnostic is applied, where the atypical observations by the one-step-ahead predictor instead of downweighting is also proposed. The performance of the proposed estimation methods is compared with other robust estimation methods by simulation study. It is observed that the iterated interpolation approach performs reasonably well is general, especially for single AO case and large $\phi$ in absolute values.

• The Korean Journal of Applied Statistics
• /
• v.1 no.2
• /
• pp.10-26
• /
• 1987

If the forecasts from different, sources are combined in some way, the resulting forecasts may be more accurate than any of the individual components. In this paper, the established procedures of combining forecasts are reviewed and the alternative procedures are suggested. By the results of empirical analysis from survey data, the method of combining forecasts using the restricted regression weights, the restricted robust regression weights, and mixed regression weights are robust. We can not find the most efficient combined forecasts in any case if we select the corresponding decision by preliminary analysis for the statistical properties of individual dorecasts, our results of combined forecast can became useful.

• Journal of the Korean Data and Information Science Society
• /
• v.26 no.3
• /
• pp.549-559
• /
• 2015

The MDPDE is an attractive alternative to maximum likelihood estimator because of the strong robustness properties that it inherently possess. The characteristics of MDPDE can be varied with the tuning parameter, in general, there is a trade-off between robustness and asymptotic efficiency. Hence, selection of optimal tuning parameter is important but complicated task. In this study, we introduce two optimal tuning parameter selection methods proposed by Fujisawa and Eguchi (2005) and Warwick (2006). Through simulation study, we found out that Warwick's method yields excessively small optimal tuning parameter in certain cases while Fujisawa and Eguchi's method performs well. Therefore, we think Fujisawa and Eguchi's method can be used commonly for finding optimal tuning parameter of MDPDE.

• The Korean Journal of Applied Statistics
• /
• v.35 no.3
• /
• pp.395-406
• /
• 2022

This paper studies the outlier detection method for multivariate long memory time series. The existing outlier detection methods are based on a short memory VARMA model, so they are not suitable for multivariate long memory time series. It is because higher order of autoregressive model is necessary to account for long memory, however, it can also induce estimation instability as the number of parameter increases. To resolve this issue, we propose outlier detection methods based on the VHAR structure. We also adapt the robust estimation method to estimate VHAR coefficients more efficiently. Our simulation results show that our proposed method performs well in detecting outliers in multivariate long memory time series. Empirical analysis with stock index shows RVHAR model finds additional outliers that existing model does not detect.

• The Korean Journal of Applied Statistics
• /
• v.20 no.3
• /
• pp.561-571
• /
• 2007

Local linear regression estimator is the most widely used nonparametric regression estimator which has a number of advantages over the traditional kernel estimators. It is well known that local linear estimator can produce erratic result in sparse regions in the realization of the design and the interpolation method of Hall and Turlach (1997) is the very efficient way to resolve this problem. However, it has been never pointed out that Hall and Turlach's interpolation method is very sensitive to outliers. In this paper, we propose the robust version of the interpolation method for adapting to sparse design. The finite sample properties of the method is compared with Hall and Turlach's method by the simulation study.

• 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.

• Journal of Korean Society of Coastal and Ocean Engineers
• /
• v.35 no.4
• /
• pp.84-94
• /
• 2023

The wave height and period regression curve is widely used to estimate the design wave period. In this study, the parameters of the curves are estimated, compared, and evaluated using the linear, robust linear, and nonlinear regression methods, respectively. The data used in the design wave height estimation are the annual maxima (AM) wave height and period data sets divided by typhoon and non-typhoon conditions, provided by the Ministry of Oceans and Fisheries (2019). The estimation parameters show significant differences in the local coastal waters and the estimation methods. The estimation parameters based on the Suh et al. (2008, 2010) method show the apparent bias, under-estimation in the intercept (scale) parameter, and over-estimation in the slope (exponent) parameter, respectively.

• The Korean Journal of Applied Statistics
• /
• v.32 no.1
• /
• pp.99-109
• /
• 2019

We compared the confidence intervals of estimators using various bootstrap methods for a Random Coefficient Autoregressive(RCA) model. We consider a Quasi score estimator and M-Quasi score estimator using Huber, Tukey, Andrew and Hempel functions as bounded functions, that do not have required assumption of distribution. A standard bootstrap method, percentile bootstrap method, studentized bootstrap method and hybrid bootstrap method were proposed for the estimations, respectively. In a simulation study, we compared the asymptotic confidence intervals of the Quasi score and M-Quasi score estimator with the bootstrap confidence intervals using the four bootstrap methods when the underlying distribution of the error term of the RCA model follows the normal distribution, the contaminated normal distribution and the double exponential distribution, respectively.

• International Journal of Highway Engineering
• /
• v.13 no.3
• /
• pp.121-129
• /
• 2011

The purpose of this study is to determine the relationship between observed traffic data and NOx concentrations from not an ideal condition but a real road in real-time. Also we aim to develop an estimation model for NOx emission concentrations due to vehicle exhaust gas, and it can be applied to monitor the degree of air pollution on National Road in real-time. To eliminate outliers which are occurred due to errors of equipments and other variables, we use the robust analysis and develop two models. which are considering and not considering wind impact. The result of this research can be used for understanding present condition of air pollution caused by vehicle exhaust gas and evaluating for environmental effects of transportation policy.

• Economic and Environmental Geology
• /
• v.28 no.4
• /
• pp.433-438
• /
• 1995

The most popular minimization method is based on the least-squares criterion, which uses the $L_2$ norm to quantify the misfit between observed and synthetic data. The solution of the least-squares problem is the maximum likelihood point of a probability density containing data with Gaussian uncertainties. The distribution of errors in the geophysical data is, however, seldom Gaussian. Using the $L_2$ norm, large and sparsely distributed errors adversely affect the solution, and the estimated model parameters may even be completely unphysical. On the other hand, the least-absolute-deviation optimization, which is based on the $L_1$ norm, has much more robust statistical properties in the presence of noise. The solution of the $L_1$ problem is the maximum likelihood point of a probability density containing data with longer-tailed errors than the Gaussian distribution. Thus, the $L_1$ norm gives more reliable estimates when a small number of large errors contaminate the data. The effect of outliers is further reduced by M-fitting method with Cauchy error criterion, which can be performed by iteratively reweighted least-squares method.