• The Korean Journal of Applied Statistics
• /
• v.6 no.2
• /
• pp.227-235
• /
• 1993

Calibration relates the estimation of independent variable which rquires more effort or expense than dependent variable does. It would be provided with high accuracy because a little change of the result of independent variable cn cause a serious effect to the human being. Usual statistical analysis assumes the normality of error distribution or linearity of data. It is desirable to analyze the data without those assumptions for the accuracy of the calibration. In this paper, we calibrated the data nonparametrically without those assumptions and derived confidence interval estimate for the independent variable. As a method, we used kernel method which is popular in modern statistical branch. We derived bootstrap confidence interval estimate from the bootstrap confidence band.

• Communications for Statistical Applications and Methods
• /
• v.26 no.2
• /
• pp.149-161
• /
• 2019

In this article, we suggest the following approaches to simultaneous variable selection and outlier detection. First, we determine possible candidates for outliers using properties of an intercept estimator in a difference-based regression model, and the information of outliers is reflected in the multiple regression model adding mean shift parameters. Second, we select the best model from the model including the outlier candidates as predictors using stochastic search variable selection. Finally, we evaluate our method using simulations and real data analysis to yield promising results. In addition, we need to develop our method to make robust estimates. We will also to the nonparametric regression model for simultaneous outlier detection and variable selection.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.339-348
• /
• 2017

One of the major issues of statistical process control for variables data is monitoring both the mean and the standard deviation. The traditional approach to monitor these parameters is to simultaneously use two seperate control charts. However there have been some works on developing a single chart using a single plotting statistic for joint monitoring, and it is claimed that they are simpler and may be more appealing than the traditonal one from a practical point of view. When using these control charts for variables data, estimating in-control parameters and checking the normality assumption are the very important step. Nonparametric Shewhart-Lepage chart, proposed by Mukherjee and Chakraborti (2012), is an attractive option, because this chart uses only a single control statistic, and does not require the in-control parameters and the underlying continuous distribution. In this paper, we introduce the Shewhart-Lepage chart, and propose the design procedure to find the optimal diagnosis limits when the location and the scale parameters change simultaneously. We also compare the efficiency of the proposed method with that of Mukherjee and Chakraborti (2012).

• The Korean Journal of Applied Statistics
• /
• v.28 no.3
• /
• pp.529-540
• /
• 2015

A cross-over design is frequently used in clinical trials (especially in bioequivalence tests with a parametric method) for the comparison of two treatments. Missing values frequently take place in cross-over designs in the second period. Usually, subjects that have missing values are removed and analyzed. However, it can be unsuitable in clinical trials with a small sample size. In this paper, we compare single imputation methods in a $2{\times}2$ cross-over design when missing values exist in the second period. Additionally, parametric and nonparametric methods are compared after applying single imputation methods. A Monte-Carlo simulation study compares type I error and the power of methods.

• Journal of the Korean Data and Information Science Society
• /
• v.28 no.2
• /
• pp.361-369
• /
• 2017

This paper analyzes the causal relationship between dollar exchange rate and international oil price. Although large literature on the relationship has accumulated, results are not unique but diversified. Based on the idea that such diversified results may be due to different causality at different economic status, we considers an approach to test the causal relationship at each quantile. This approach is different from the mean causality analysis widely employed by the existing literature of the causal relationship. In this paper, monthly data from May 1987 to 2013 is used for the causal analysis in which Brent oil price and Major Currencies Dollar Index (MCDI) are considered. The test method is the nonparametric test for causality in quantile suggested by Jeong et al. (2012). The results show that although dollar exchange rate causes oil price in mean, the causal relationship does not exist at most quantiles.

• The Korean Journal of Applied Statistics
• /
• v.26 no.2
• /
• pp.291-305
• /
• 2013

Collective traffic data (BPS, PPS etc.) for detection against the distributed denial of service attack on network is the time sequencing big data. The algorithm to detect the change point in the big data should be accurate and exceed in detection time and detection capability. In this work, the sliding window and discretization method is used to detect the change point in the big data, and propose five nonparametric test statistics using empirical distribution functions and ranks. With various distribution functions and their parameters, the detection time and capability including the detection delay time and the detection ratio for five test methods are explored and discussed via monte carlo simulation and illustrative examples.

• The Korean Journal of Applied Statistics
• /
• v.13 no.2
• /
• pp.447-455
• /
• 2000

We propose a weighted kernel pattern recognition method using the K -means clustering algorithm to reduce computation and storage required for the full kernel classifier. This technique finds a set of reference vectors and weights which are used to approximate the kernel classifier. Since the hierarchical clustering method implemented in the 'Weighted Parzen Window (WP\V) classifier is not able to rearrange the proper clusters, we adopt the K -means algorithm to find reference vectors and weights from the more properly rearranged clusters \Ve find that the proposed method outperforms the \VP\V method for the repre~entativeness of the reference vectors and the data reduction.

• The Korean Journal of Applied Statistics
• /
• v.19 no.3
• /
• pp.587-598
• /
• 2006

The number of variables exceeds the number of samples in microarray data. We propose a nonparametric local linear logistic classification procedure using orthogonal components for classifying high-dimensional microarray data. The proposed method is based on the local likelihood and can be applied to multi-class classification. We applied the local linear logistic classification method using PCA, PLS, and factor analysis components as new features to Leukemia data and colon data, and compare the performance of the proposed method with the conventional statistical classification procedures. The proposed method outperforms the conventional ones for each component, and PLS has shown best performance when it is embedded in the proposed method among the three orthogonal components.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.7 no.4
• /
• pp.598-603
• /
• 2003

This paper presents a sub-band feature extraction approach in which the feature extraction method in the individual frequency sub-bands is determined in terms of speech recognition accuracy. As in the multi-band paradigm, features are extracted independently in frequency sub-regions of the speech signal. Since the spectral shape is well structured in the low frequency region, the all pole model is effective for feature extraction. But, in the high frequency region, the nonparametric transform, discrete cosine transform is effective for the extraction of cepstrum. Using the sub-band specific feature extraction method, the linguistic information in the individual frequency sub-bands can be extracted effectively for automatic speech recognition. The validity of the proposed method is shown by comparing the results of speech recognition experiments for our method with those obtained using a full-band feature extraction method.

• Communications for Statistical Applications and Methods
• /
• v.17 no.2
• /
• pp.141-151
• /
• 2010

Hybrid fuzzy regression analysis is used for integrating randomness and fuzziness into a regression model. Least squares support vector machine(LS-SVM) has been very successful in pattern recognition and function estimation problems for crisp data. This paper proposes a new method to evaluate hybrid fuzzy linear and nonlinear regression models with crisp inputs and fuzzy output using weighted fuzzy arithmetic(WFA) and LS-SVM. LS-SVM allows us to perform fuzzy nonlinear regression analysis by constructing a fuzzy linear regression function in a high dimensional feature space. The proposed method is not computationally expensive since its solution is obtained from a simple linear equation system. In particular, this method is a very attractive approach to modeling nonlinear data, and is nonparametric method in the sense that we do not have to assume the underlying model function for fuzzy nonlinear regression model with crisp inputs and fuzzy output. Experimental results are then presented which indicate the performance of this method.