• 응용통계연구
• /
• 제5권1호
• /
• pp.103-111
• /
• 1992

다변량 통계분석기법중 하나로 제기된 투사지향방법은 다변량자료를 관심있는 일차원 또는 이차원의 자료로의 선형투사를 찾아 나가는 방법이다. 이 방법은 다변량 자료가 갖는 차원의 문제를 해결해 줄 수 있는 유용한 기법으로 제시되었다. 본 연구에서는 투사지향방법을 이용하여 추정한 다변량 확률밀도함수를 사용한 새로운 비모수적인 판별분석방법을 제시하고, 이를 기존의 모수적 판별분석방법중 실제적으로 많이 사용되는 선형판별함수방법, 그리고 기존의 비모수적 판별분석방법중 계산상의 편리성이 많은 K-최인접방법과 컴퓨터 시뮬레이션을 통하여 비교분석하였다.

• Communications for Statistical Applications and Methods
• /
• 제7권1호
• /
• pp.225-232
• /
• 2000

The linear and quadratic discrimination functions based on normal theory are widely used to classify an observation to one of predefined groups. But the discriminant functions are sensitive to outliers. A high breakdown procedure to estimate location and scatter of multivariate data is the minimum volume ellipsoid or MVE estimator To obtain high breakdown classifiers outliers in multivariate data are detected by using the robust Mahalanobis distance based on MVE estimators and the weighted estimators are inserted in the functions for classification. A samll-sample MOnte Carlo study shows that the high breakdown robust procedures perform better than the classical classifiers.

• 응용통계연구
• /
• 제7권2호
• /
• pp.145-157
• /
• 1994

확률적 순서관계를 갖는 다변량분포에서 얻어진 자료가 결측값을 갖는 불완전한 자료일 때, EM 알고리즘을 이용한 최우추정법을 논의하였다. 본 논문에서는 관찰값들이 부분적으로 분류된 분할표자료에 국한하여 연구되었으며 기존의 동위회귀추정 프로그램을 써서 EM을 수행할 수 있는 이점이 있다. 예를 통하여 제안된 추정법을 설명한다.

• Communications for Statistical Applications and Methods
• /
• 제20권1호
• /
• pp.23-28
• /
• 2013

We propose a version of Cook's distance (called local distance) in the multivariate linear model. The proposed version is a matrix, while the existing version of Cook's distance (called global distance) is a scalar. The existing Cook's distance is the trace of the proposed Cook's distance. In addition, we argue that the proposed Cook's distance has a more natural extension of the Cook's distance in the univariate linear model than the existing Cook's distance. An illustrative example based on a real data set is given.

• International Journal of Reliability and Applications
• /
• 제9권1호
• /
• pp.7-15
• /
• 2008

The dimensional measurement equipment, such as Coordinate Measurement Machine (CMM), Optical Coordinate Measurement Machine (OCMM), and Checking Fixture (CF), take multiple dimensional measurements for each part in an automobile industry. Measurements are also recorded under different measurement systems to see if the responses differ significantly over these systems. Each measurement system (CMM, OCMM, and CF) will be considered as different treatments. This set-up provides massive amounts of process data which are multivariate in nature. Therefore, the multivariate statistical analysis is required to analyze data that are dependent on each other. This research provides step by step methodology for the evaluation procedure of the compatibility of measurement systems and clarify a systematic analyzation among the different measurement system's compatibility followed by number of case studies for each methodologies provided.

• Journal of the Korean Data and Information Science Society
• /
• 제17권3호
• /
• pp.973-982
• /
• 2006

Performances of variable sampling interval(VSI) multivariate control charts with accumulate-combine approach for monitoring mean vector of p related quality variables were investigated. Shewhart control chart is also proposed to compare the performances of CUSUM and EWMA charts. Numerical comparisons show that performances of CUSUM and EWMA charts are more efficient than Shewhart chart for small or moderate shifts, and VSI chart is more efficient than fixed sampling interval(FSI) chart. We also found that performances of the CUSUM or EWMA chart with accumulate-combine approach are substantially efficient than those of Shewhart chart.

• Journal of the Korean Data and Information Science Society
• /
• 제28권4호
• /
• pp.911-925
• /
• 2017

We investigate the properties of multivariate Shewart control charts with VSI procedure for monitoring simultaneous monitoring mean vector and covariance matrix in term of ANSW (average number of switches), probability of switch and ASI (average sampling interval), ATS (average time to signal). From examining the ANSW values, we know that it does not switch frequently. The VSI control charts are superior to the corresponding FSI control charts in terms of ATS. And, it can be also seen that the VSI procedures have substantially fewer switches for small or moderate shifts of the mean vector and variances.

• Communications for Statistical Applications and Methods
• /
• 제25권1호
• /
• pp.71-78
• /
• 2018

In this study, we consider the likelihood ratio test for the covariance matrix of the multivariate normal data. For this, we propose a method for obtaining null distributions of the likelihood ratio statistics by the Monte-Carlo approach when it is difficult to derive the exact null distributions theoretically. Then we compare the performance and precision of distributions obtained by the asymptotic normality and the Monte-Carlo method for the likelihood ratio test through a simulation study. Finally we discuss some interesting features related to the likelihood ratio test for the covariance matrix and the Monte-Carlo method for obtaining null distributions for the likelihood ratio statistics.

• Quantitative Bio-Science
• /
• 제37권2호
• /
• pp.73-79
• /
• 2018

As high frequency (HF, for short) time series is now prevalent in the presence of real time big data, volatility computations based on traditional ARCH/GARCH models need to be further developed to suit the high frequency characteristics. This article reviews realized volatilities (RV) and multivariate GARCH (MGARCH) to deal with high frequency volatility computations. As a (functional) infinite dimensional models, the fARCH and fGARCH are introduced to accommodate ultra high frequency (UHF) volatilities. The fARCH and fGARCH models are developed in the recent literature by Hormann et al. [1] and Aue et al. [2], respectively, and our discussions are mainly based on these two key articles. Real data applications to domestic UHF financial time series are illustrated.

• Communications for Statistical Applications and Methods
• /
• 제28권3호
• /
• pp.267-279
• /
• 2021

A regression with multi-dimensional responses is quite common nowadays in the so-called big data era. In such regression, to relieve the curse of dimension due to high-dimension of responses, the dimension reduction of predictors is essential in analysis. Sufficient dimension reduction provides effective tools for the reduction, but there are few sufficient dimension reduction methodologies for multivariate regression. To fill this gap, we newly propose two fused slice-based inverse regression methods. The proposed approaches are robust to the numbers of clusters or slices and improve the estimation results over existing methods by fusing many kernel matrices. Numerical studies are presented and are compared with existing methods. Real data analysis confirms practical usefulness of the proposed methods.