• The Korean Journal of Applied Statistics
• /
• v.29 no.7
• /
• pp.1155-1163
• /
• 2016

This paper deals with an estimation procedure of variance components in a mixed effects model by projections. Projections are used to obtain sums of squares instead of using reductions in sums of squares due to fitting both the assumed model and sub-models in the fitting constants method. A projection matrix can be obtained for the residual model at each step by a stepwise procedure to test the hypotheses. A weighted least squares method is used for the estimation of fixed effects. Satterthwaite's approximation is done for the confidence intervals for variance components.

• Water for future
• /
• v.28 no.5
• /
• pp.219-233
• /
• 1995

A non-stationary multivariate model is selected in which the mean and variance of rainfall are not temporally or spatially constant. And the rainfall prediction system is constructed which uses the recursive estimation algorithm, Kalman filter, to estimate system states and parameters of rainfall model simulataneously. The on-line, real-time, multivariate short-term, rainfall prediction for multi-stations and lead-times is carried out through the estimation of non-stationary mean and variance by the storm counter method, the normalized residual covariance and rainfall speed. The results of rainfall prediction system model agree with those generated by non-stationary multivariate model. The longer the lead time is, the larger the root mean square error becomes and the further the model efficiency decreases form 1. Thus, the accuracy of the rainfall prediction decreases as the lead time gets longer. Also it shows that the mean obtained by storm counter method constitutes the most significant part of the rainfall structure.

• The Korean Journal of Applied Statistics
• /
• v.37 no.1
• /
• pp.39-48
• /
• 2024

In this paper, we introduce an extended GARCH model designed to address asymmetric leverage effects. The variance in the standard GARCH model is composed of past conditional variances and past squared residuals. However, it is not possible to model asymmetric leverage effects with squared residuals alone, so in this paper, we propose a new extended GARCH model to explain the leverage effects using the Yeo-Johnson transformation which adjusts transformation parameter to make asymmetric data more normal or symmetric. We utilize the reverse properties of Yeo-Johnson transformation to model asymmetric volatility. We investigate the characteristics of the proposed model and parameter estimation. We also explore how to derive forecasts and forecast intervals in the proposed model. We compare it with standard GARCH and other extended GARCH models that model asymmetric leverage effects through empirical data analysis.

• 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.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.9 no.1
• /
• pp.47-54
• /
• 1991

In this study, the accuracy of three dimensional location was improved by self calibration bundle method with additional parameter, which is to correct systematic error through detection and elimination of the gross error from updated reference variance for observation values in photogram-metry. In this study, with the result of comparing accuracy of each method, correcting systematic error is more effective after gross error detection and when observation values are contained more than two gross error the point with maximum correlation value is detected by masking effect of least square adjustment.

• Journal of the Korean Data and Information Science Society
• /
• v.16 no.3
• /
• pp.591-601
• /
• 2005

In this paper we consider detection of a discontinuity point in the variance function. When the mean function is discontinuous at a point, the variance function is usually discontinuous at the point. In this case, we had better estimate the location of the discontinuity point with the mean function rather than the variance function. On the other hand, the variance function only has a discontinuity point. The target function in order to estimate the location can be used the second moment function since the variance function and the second moment function have the same location and jump size of the discontinuity point. We propose a nonparametric detection method of the discontinuity point with the second moment function. We give the asymptotic results of these estimators. Computer simulation demonstrates the improved performance of the method over the existing ones.

• Journal of Korea Water Resources Association
• /
• v.39 no.12 s.173
• /
• pp.997-1012
• /
• 2006

We have presented a nonparametric stochastic approach for the SOI(Southern Oscillation Index) series that used nonlinear methodology called Nonlinear AutoRegressive(NAR) based on conditional kernel density function and CAFPE(Corrected Asymptotic Final Prediction Error) lag selection. The fitted linear AR model represents heteroscedasticity, and besides, a BDS(Brock - Dechert - Sheinkman) statistics is rejected. Hence, we applied NAR model to the SOI series. We can identify the lags 1, 2 and 4 are appropriate one, and estimated conditional mean function. There is no autocorrelation of residuals in the Portmanteau Test. However, the null hypothesis of normality and no heteroscedasticity is rejected in the Jarque-Bera Test and ARCH-LM Test, respectively. Moreover, the lag selection for conditional standard deviation function with CAFPE provides lags 3, 8 and 9. As the results of conditional standard deviation analysis, all I.I.D assumptions of the residuals are accepted. Particularly, the BDS statistics is accepted at the 95% and 99% significance level. Finally, we split the SOI set into a sample for estimating themodel and a sample for out-of-sample prediction, that is, we conduct the one-step ahead forecasts for the last 97 values (15%). The NAR model shows a MSEP of 0.5464 that is 7% lower than those of the linear model. Hence, the relevance of the NAR model may be proved in these results, and the nonparametric NAR model is encouraging rather than a linear one to reflect the nonlinearity of SOI series.

• The Korean Journal of Applied Statistics
• /
• v.30 no.3
• /
• pp.335-344
• /
• 2017

This paper discusses a method of analyzing data from split-plot experiments by projections. The assumed model for data has two experimental errors due to two different experimental sizes and some random components in treatment effects. Residual random models are constructed to obtain sums of squares due to random effects. Expectations of sums of squares are obtained by Hartley's synthesis. Estimable functions of fixed effects are discussed.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 2003.05a
• /
• pp.151-157
• /
• 2003

OLS로 알려진 기존의 주절 방법은 변수수의 증가에 따라 다중공선성(Multicollinearity)의 문제와 더불어 해석력(Interpretability)이 떨어지는 문제를 가지게 된다. 본 연구에서는 파라미터의 절대값의 크기(L1-Norm)에 제약을 줌으로써 이와 같은 OLS의 문제를 해결할 수 있는 동시에, 잔차의 제곱합대신 절대오차를 사용하는 Least Absolute Value(LAV) 방법을 사용함으로써 이상치에 로버스트한 결과를 주는 방법론을 제안한다. 또한. 본 연구에서 제안하는 방법이 선형계획법에 의해 모델처럼 될 수 있는 특성으로 인해 제약조건이 있는 이차 형태의 최적화 문제보다 수행 속도면에서 뛰어난 결과를 주는 것을 수치예제을 통해 보인다.

• Journal of the Korean Data and Information Science Society
• /
• v.3 no.1
• /
• pp.1-16
• /
• 1992

In regression diagnostics, a number of joint influence measures based on various statistical tools have been discussed. We consider an alternate representation in terms of the predicted residual and g-leverage determined by the remaining points. By this approach, we choose the predicted residual sum of squares for the keypoints as joint influence measure and propose a new expression of it so that we can extend the single case form to the multiple case one. Furthermore we suggest a seach method for joint influence after investigating some properties of the new expression.