• Communications for Statistical Applications and Methods
• /
• v.13 no.2
• /
• pp.233-241
• /
• 2006

Classical principal component analysis (PCA) can be formulated as finding the linear subspace that best accommodates multidimensional data points in the sense that the sum of squared residual distances is minimized. As alternatives to such LS (least squares) fitting approach, we produce LMS (least median of squares) and LTS (least trimmed squares)-type PCA by minimizing the median of squared residual distances and the trimmed sum of squares, in a similar fashion to Rousseeuw (1984)'s alternative approaches to LS linear regression. Proposed methods adopt the data-driven optimization algorithm of Croux and Ruiz-Gazen (1996, 2005) that is conceptually simple and computationally practical. Numerical examples are given.

• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.89-100
• /
• 1994

In this paper we consider a new estimator of mean residual life (MRL), based on the partial moment of the distribution. The parameters of a partial moment are estimated by its maximum likelihood estimators when the underlying distribution is known. Though the new estimator is not a consistent estimator of the MRL, it is shown to have smaller mean squared error than the well known empirical MRL estimator for certain parametric families. Numerical summaries of the mean squared errors of the new estimator are presented.

• Journal of Wetlands Research
• /
• v.21 no.spc
• /
• pp.90-97
• /
• 2019

In recent years, flood due to the consecutive storm events have been occurred and property damage and casualties are in increasing trend. This study calls the consecutively occurred storm events as a mega rainfall scenario and the discharge by the scenario is defined as a mega flood discharge. A mega rainfall scenario was created on the assumption that 100-year frequency rainfall events were consecutively occurred in the Gyeongancheon stream basin. The SSARR (Streamflow Synthesis and Reservoir Regulation) model was used to estimate the mega flood discharge using the scenario in the basin. In addition, in order to perform more reasonable runoff analysis, the parameters were estimated using the SCE_UA algorithm. Also, the calibration and verification were performed using the objective functions of the weighted sum of squared of residual(WSSR), which is advantageous for the peak discharge simulation and sum of squared of residual(SSR). As a result, the mega flood discharge due to the continuous occurrence of 100-year frequency rainfall events in the Gyeongan Stream Basin was estimated to be 4,802㎥/s, and the flood discharge due to the 100-year frequency single rainfall event estimated by "the Master Plan for the Gyeongancheon Stream Improvement" (2011) was 3,810㎥/s. Therefore, the mega flood discharge was found to increase about 992㎥/s more than the single flood event. The results of this study can be used as a basic data for Comprehensive Flood Control Plan of the Gyeongan Stream basin.

• Ocean Science Journal
• /
• v.43 no.2
• /
• pp.81-90
• /
• 2008

A study of estuarine flows during a neap tide was performed using 13-hour roving acoustic Doppler current profiles (ADCP) and conductivity-temperature-depth (CTD) profiles in the Altamaha River estuary, Georgia, U.S.A. The least-squared harmonic analysis method was used to fit the tidal ($M_2$) component and separate the flow into two components: the tidal and residual ($M_2$-removed) flows. We applied this method to depth-averaged data. Results show that the $M_2$ component demonstrates over 95% of the variability of observation data. As the flow was dominated by the $M_2$ tidal component in a narrow channel, the tidal ellipse distribution was essentially a back-and-forth motion. The amplitude of $M_2$ velocity component increased slightly from the river mouth (0.45 m/sec) to land (0.6 m/sec) and the phase showed fairly constant values in the center of the channel and rapidly decreasing values near the northern and southern shoaling areas. The residual flow and transport calculated from depth-averaged flow shows temporal variability over the tidal time scale. Strong landward flows appeared during slack waters which may be attributed to increased baroclinic forcing when turbulent mixing decreases.

• Journal of the Korean Society of Radiology
• /
• v.16 no.3
• /
• pp.295-302
• /
• 2022

Sparse-view computed tomography (CT) imaging technique is able to reduce radiation dose, ensure the uniformity of image characteristics among projections and suppress noise. However, the reconstructed images obtained by the sparse-view CT imaging technique suffer from severe artifacts, resulting in the distortion of image quality and internal structures. In this study, we proposed a convolutional neural network (CNN) with wavelet transformation and residual learning for reducing artifacts in sparse-view CT image, and the performance of the trained model was quantitatively analyzed. The CNN consisted of wavelet transformation, convolutional and inverse wavelet transformation layers, and input and output images were configured as sparse-view CT images and residual images, respectively. For training the CNN, the loss function was calculated by using mean squared error (MSE), and the Adam function was used as an optimizer. Result images were obtained by subtracting the residual images, which were predicted by the trained model, from sparse-view CT images. The quantitative accuracy of the result images were measured in terms of peak signal-to-noise ratio (PSNR) and structural similarity (SSIM). The results showed that the trained model is able to improve the spatial resolution of the result images as well as reduce artifacts in sparse-view CT images effectively. Also, the trained model increased the PSNR and SSIM by 8.18% and 19.71% in comparison to the imaging model trained without wavelet transformation and residual learning, respectively. Therefore, the imaging model proposed in this study can restore the image quality of sparse-view CT image by reducing artifacts, improving spatial resolution and quantitative accuracy.

• Communications for Statistical Applications and Methods
• /
• v.12 no.3
• /
• pp.557-587
• /
• 2005

This paper elucidates the limiting Gaussian distribution of a class of rank order statistics {$T_N$} for two sample problem pertaining to empirical processes of the squared residuals from two independent samples of GARCH processes. A distinctive feature is that, unlike the residuals of ARMA processes, the asymptotics of {$T_N$} depend on those of GARCH volatility estimators. Based on the asymptotics of {$T_N$}, we empirically assess the relative asymptotic efficiency and effect of the GARCH specification for some GARCH residual distributions. In contrast with the independent, identically distributed or ARMA settings, these studies illuminate some interesting features of GARCH residuals.

• Journal of the Korean Statistical Society
• /
• v.34 no.2
• /
• pp.161-172
• /
• 2005

In this paper a nonparametric method is proposed for detecting conditionally heteroscedastic errors in a nonparametric time series regression model where the observation points are equally spaced on [0,1]. It turns out that the first-order sample autocorrelation of the squared residuals from the kernel regression estimates provides essential information. Illustrative simulation study is presented for diverse errors such as ARCH(1), GARCH(1,1) and threshold-ARCH(1) models.

• ETRI Journal
• /
• v.42 no.3
• /
• pp.333-340
• /
• 2020

We consider a hybrid combiner design for downlink massive multiple-input multiple-output systems when there is residual inter-user interference and each user is equipped with a limited number of radio frequency (RF) chains (less than the number of receive antennas). We propose a hybrid combiner that minimizes the mean-squared error (MSE) between the information symbols and the ones estimated with a constant amplitude constraint on the RF combiner. In the proposed scheme, an iterative alternating optimization method is utilized. At each iteration, one of the analog RF and digital baseband combining matrices is updated to minimize the MSE by fixing the other matrix without considering the constant amplitude constraint. Then, the other matrix is updated by changing the roles of the two matrices. Each element in the RF combining matrix is obtained from the phase component of the solution matrix of the optimization problem for the RF combining matrix. Simulation results show that the proposed scheme performs better than conventional matrix-decomposition schemes.

• Steel and Composite Structures
• /
• v.36 no.2
• /
• pp.143-161
• /
• 2020

In nanosized structures as the surface area to the bulk volume ratio increases the classical continuum mechanics approaches fails to investigate the mechanical behavior of such structures. In perforated nanobeam structures, more decrease in the bulk volume is obtained due to perforation process thus nonclassical continuum approaches should be employed for reliable investigation of the mechanical behavior these structures. This article introduces an analytical methodology to investigate the size dependent, surface energy, and perforation impacts on the nonclassical bending behavior of regularly squared cutout nanobeam structures for the first time. To do this, geometrical model for both bulk and surface characteristics is developed for regularly squared perforated nanobeams. Based on the proposed geometrical model, the nonclassical Gurtin-Murdoch surface elasticity model is adopted and modified to incorporate the surface energy effects in perforated nanobeams. To investigate the effect of shear deformation associated with cutout process, both Euler-Bernoulli and Timoshenko beams theories are developed. Mathematical model for perforated nanobeam structure including surface energy effects are derived in comprehensive procedure and nonclassical boundary conditions are presented. Closed forms for the nonclassical bending and rotational displacements are derived for both theories considering all classical and nonclassical kinematics and kinetics boundary conditions. Additionally, both uniformly distributed and concentrated loads are considered. The developed methodology is verified and compared with the available results and an excellent agreement is noticed. Both classical and nonclassical bending profiles for both thin and thick perforated nanobeams are investigated. Numerical results are obtained to illustrate effects of beam filling ratio, the number of hole rows through the cross section, surface material characteristics, beam slenderness ratio as well as the boundary and loading conditions on the non-classical bending behavior of perforated nanobeams in the presence of surface effects. It is found that, the surface residual stress has more significant effect on the bending deflection compared with the corresponding effect of the surface elasticity, Es. The obtained results are supportive for the design, analysis and manufacturing of perforated nanobeams.

• Communications for Statistical Applications and Methods
• /
• v.18 no.6
• /
• pp.733-739
• /
• 2011

The linear absolute shrinkage and selection operator(Lasso) method improves the low prediction accuracy and poor interpretation of the ordinary least squares(OLS) estimate through the use of $L_1$ regularization on the regression coefficients. However, the Lasso is not robust to outliers, because the Lasso method minimizes the sum of squared residual errors. Even though the least absolute deviation(LAD) estimator is an alternative to the OLS estimate, it is sensitive to leverage points. We propose a robust Lasso estimator that is not sensitive to outliers, heavy-tailed errors or leverage points.