• Communications for Statistical Applications and Methods
• /
• v.26 no.4
• /
• pp.371-383
• /
• 2019

Panel data sets have been developed in various areas, and many recent studies have analyzed panel, or longitudinal data sets. Maximum likelihood (ML) may be the most common statistical method for analyzing panel data models; however, the inference based on the ML estimate will have an inflated Type I error because the ML method tends to give a downwardly biased estimate of variance components when the sample size is small. The under estimation could be severe when data is incomplete. This paper proposes the restricted maximum likelihood (REML) method for a random effects panel data model with a censored dependent variable. Note that the likelihood function of the model is complex in that it includes a multidimensional integral. Many authors proposed to use integral approximation methods for the computation of likelihood function; however, it is well known that integral approximation methods are inadequate for high dimensional integrals in practice. This paper introduces to use the moments of truncated multivariate normal random vector for the calculation of multidimensional integral. In addition, a proper asymptotic standard error of REML estimate is given.

• Journal of Sasang Constitutional Medicine
• /
• v.18 no.3
• /
• pp.131-144
• /
• 2006

1. Objectives This research uses the DNA chip, which includes 16,383 gene code, and various statistic prediction way that shows objectification index for the objectification of constitution diagnosis. 2. Methods Drawing blood whose constitution is confirmed, and analyze its gene information by using 1.7k DNA chip to find the gene correlation through multivariate statistical method. 3. Results and Conclusions Distinctive genes such as AK001919, U09384, NM_001805, X99962, NM_004796, AK026738, AL050148, BC002538, AK027074, AK026219, AF087962, AL390142, NM_015372, AL157466, NM_002446, AK024523, NM_014706, NM_014746 and AL137544 were related to Taeumin; AL157448, NM_005957, NM_005656, NM_017548, AK027246, NM_003025, NM_012302 and NM_005905 were represented in Soeumin, while AK026503, AF147325, NM_002076, AF147307, AK001375, NM_003740, NM_005114, AB007890, NM_005505, NM_015900, NM_014936, Z70694, AB023154, U52076, NM_004360, NM_005835, NM_017528, AF087987, NM_014897, AK021720, NM_006420, AJ277915, AK002118 and AK021918 were for Soyangin. This study figured out the possibility to develop the prediction system by sorting each constitution's gene, and research each constitution's distinctive character of manifestation pattern.

• The Korean Journal of Applied Statistics
• /
• v.35 no.5
• /
• pp.579-591
• /
• 2022

Functional data is collected in various fields. It is often necessary to test whether there are differences among groups of functional data. In this case, it is not appropriate to explain using the point-wise ANOVA method, and we should present not the point-wise result but the integrated result. Various studies on functional data analysis of variance have been proposed, and recently implemented those methods in the package fdANOVA of R. In this paper, I first explain ANOVA and multivariate ANOVA, then I will introduce various methods of analysis of variance for univariate and multivariate functional data recently proposed. I also describe how to use the R package fdANOVA. This package is used to test equality of weekly temperatures in Seoul and Busan through univariate functional data ANOVA, and to test equality of multivariate functional data corresponding to handwritten images using multivariate function data ANOVA.

• The Korean Journal of Applied Statistics
• /
• v.28 no.3
• /
• pp.541-559
• /
• 2015

VaR is now widely used as an important tool to evaluate and manage financial risks. In particular, it is important to select an appropriate volatility model for the rate of return of financial assets. In this study, both univariate and multivariate models are considered to evaluate VaR of the portfolio composed of KOSPI, Hang-Seng, Nikkei indexes, and their performances are compared through back testing techniques. Overall, multivariate models are shown to be more appropriate than univariate models to estimate the portfolio VaR, in particular DCC and ADCC models are shown to be more superior than others.

• Journal of Korean Society of Forest Science
• /
• v.105 no.2
• /
• pp.157-166
• /
• 2016

This study was carried out to classify the current forest cover types and to propose the successional trends in the natural forest of Mt. Deogyu. The vegetation data were collected by the point-centered quarter method. The forest cover types were classified by various multivariate statistical analysis methods such as cluster analysis, indicator species analysis and multiple discriminant analysis. This forests were classified into five forest types by the species composition of upper layer and topographic positions: Quercus mongolica forest in the ridge, Fraxinus mandushurica-F. rhynchophylla-Cornus controversa forest and F. mandushurica forest in the valley, the Q. serrata - Pinus densiflora - Q. mongolica forest and P. densiflora forest in the low-slope. As a result of the forest successional trends depending on ecological and environmental characteristics in each forest type, the current forest types were expected that the forest succession would be proceeded toward Q. mongolica forest, F. mandshurica forest, mixed mesophytic forest, and oak-Carpinus laxiflora forest.

• The Korean Journal of Applied Statistics
• /
• v.22 no.5
• /
• pp.995-1005
• /
• 2009

Conditional correlation between financial time series plays an important role in risk management, asset allocation and portfolio selection and therefore diverse efforts for modeling conditional correlations in multivariate-GARCH processes have been made in last two decades. In particular, CCC (cf. Bollerslev, 1990) and DCC(dynamic conditional correlation, cf. Engle, 2002) models have been commonly used since they are relatively parsimonious in the number of parameters involved. This article is concerned with DCC modeling for multivariate GARCH processes in comparison with CCC specification. Various multivariate financial time series are analysed to illustrate possible advantages of DCC over CCC modeling.

• 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.

• Journal of Korean Society of Transportation
• /
• v.33 no.4
• /
• pp.368-378
• /
• 2015

This study aims to analyse the differences of job satisfaction in road freight transportation industry workers by different types of employment. The researchers utilized reliability test and factor analysis to estimate the validity and feasibility of the questionnaire. Multivariate analysis of variance (MANOVA) was also applied to assess the differences of job satisfaction level by different employment types. The results of reliability test and factor analysis clearly show that questionnaire samples are reliable and feasible. The multivariate analysis of variance result shows statistical insignificance in the level of job satisfaction between part-time workers and special type ones. On the other hand, there was a significant difference between full-time workers and those in other types of employment. The significant variables such as income, welfare, and working hour, etc were discovered.

• Communications for Statistical Applications and Methods
• /
• v.4 no.3
• /
• pp.637-644
• /
• 1997

Sometimes, the first principal component may come logically from the established knowledge and premises. For example, for the high school students' test scores of Korean, English, Mathematics, Social Study, and Science, it is natural to define the first principal component as the average of all subject scores. In such cases, we need to respect both the background knowledge and the data exploration. The aim of this study is to find the remaining components in principal component analysis of multivariate data when the first principal component is defined a priori by the researcher. Moreover, we study related matrix decomposition and their application to the graphical display.

• Journal of Korean Society for Atmospheric Environment
• /
• v.9 no.3
• /
• pp.207-215
• /
• 1993

In order to statistically predict $O_3$ levels in Seoul, the study used the TMS (telemeted air monitoring system) data from the Department of Environment, which have monitored at 20 sites in 1989 and 1990. Each data in each site was characterized by 6 major criteria pollutants ($SO_2, TSP, CO, NO_2, THC, and O_3$) and 2 meteorological parameters, such as wind speed and wind direction. To select proper variables and to determine each pollutant's behavior, univariate statistical analyses were extensively studied in the beginning, and then various applied statistical techniques like cluster analysis, regression analysis, and expert system have been intensively examined. For the initial study of high level $O_3$ prediction, the raw data set in each site was separated into 2 group based on 60 ppb $O_3$ level. A hierarchical cluster analysis was applied to classify the group based on 60 ppb $O_3$ into small calsses. Each class in each site has its own pattern. Next, multiple regression for each class was repeatedly applied to determine an $O_3$ prediction submodel and to determine outliers in each class based on a certain level of standardized redisual. Thus, a prediction submodel for each homogeneous class could be obtained. The study was extended to model $O_3$ prediction for both on-time basis and 1-hr after basis. Finally, an expect system was used to build a unified classification rule based on examples of the homogenous classes for all of sites. Thus, a concept of high level $O_3$ prediction model was developed for one of $O_3$ alert systems.