Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
Korean Journal of Applied Statistics
Journal Basic Information
Journal DOI :
The Korean Statistical Society
Editor in Chief :
Volume & Issues
Volume 26, Issue 6 - Dec 2013
Volume 26, Issue 5 - Oct 2013
Volume 26, Issue 4 - Aug 2013
Volume 26, Issue 3 - Jun 2013
Volume 26, Issue 2 - Apr 2013
Volume 26, Issue 1 - Feb 2013
Selecting the target year
Comparison of GEE Estimation Methods for Repeated Binary Data with Time-Varying Covariates on Different Missing Mechanisms
Park, Boram ; Jung, Inkyung ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 697~712
DOI : 10.5351/KJAS.2013.26.5.697
When analyzing repeated binary data, the generalized estimating equations(GEE) approach produces consistent estimates for regression parameters even if an incorrect working correlation matrix is used. However, time-varying covariates experience larger changes in coefficients than time-invariant covariates across various working correlation structures for finite samples. In addition, the GEE approach may give biased estimates under missing at random(MAR). Weighted estimating equations and multiple imputation methods have been proposed to reduce biases in parameter estimates under MAR. This article studies if the two methods produce robust estimates across various working correlation structures for longitudinal binary data with time-varying covariates under different missing mechanisms. Through simulation, we observe that time-varying covariates have greater differences in parameter estimates across different working correlation structures than time-invariant covariates. The multiple imputation method produces more robust estimates under any working correlation structure and smaller biases compared to the other two methods.
Objective Bayesian Estimation of Two-Parameter Pareto Distribution
Son, Young Sook ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 713~723
DOI : 10.5351/KJAS.2013.26.5.713
An objective Bayesian estimation procedure of the two-parameter Pareto distribution is presented under the reference prior and the noninformative prior. Bayesian estimators are obtained by Gibbs sampling. The steps to generate parameters in the Gibbs sampler are from the shape parameter of the gamma distribution and then the scale parameter by the adaptive rejection sampling algorism. A numerical study shows that the proposed objective Bayesian estimation outperforms other estimations in simulated bias and mean squared error.
Actuarial Analyses of Long Term Care Insurance for the Elderly in Korea
Kwon, Hyuk-Sung ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 725~736
DOI : 10.5351/KJAS.2013.26.5.725
Retirement income is an important personal and social issue. Problems associated with financial risk wil1 become more pronounced with the growth in the elderly population. Medical expenses in senescence is closely related to financial risk; in addition, some diseases that require long term care will increase financial risk which result in lower quality of life for the elderly. Therefore, it is necessary to understand expected long-term care costs and to manage financial risk from the perspective of an individual. This study evaluate the length of period in which a person is expected to need long term care and actuarial present values of the total cost which needs to be prepared for the care through the Korean public long term care system based on the experience data obtained from Long Term Care Insurance for the Elderly in Korea and a multi-state model.
Bayesian Interval Estimation of Tobit Regression Model
Lee, Seung-Chun ; Choi, Byung Su ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 737~746
DOI : 10.5351/KJAS.2013.26.5.737
The Bayesian method can be applied successfully to the estimation of the censored regression model introduced by Tobin (1958). The Bayes estimates show improvements over the maximum likelihood estimate; however, the performance of the Bayesian interval estimation is questionable. In Bayesian paradigm, the prior distribution usually reflects personal beliefs about the parameters. Such subjective priors will typically yield interval estimators with poor frequentist properties; however, an objective noninformative often yields a Bayesian procedure with good frequentist properties. We examine the performance of frequentist properties of noninformative priors for the Tobit regression model.
A Comparative Study of Covariance Matrix Estimators in High-Dimensional Data
Lee, DongHyuk ; Lee, Jae Won ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 747~758
DOI : 10.5351/KJAS.2013.26.5.747
The covariance matrix is important in multivariate statistical analysis and a sample covariance matrix is used as an estimator of the covariance matrix. High dimensional data has a larger dimension than the sample size; therefore, the sample covariance matrix may not be suitable since it is known to perform poorly and event not invertible. A number of covariance matrix estimators have been recently proposed with three different approaches of shrinkage, thresholding, and modified Cholesky decomposition. We compare the performance of these newly proposed estimators in various situations.
On Tail Probabilities of Continuous Probability Distributions with Heavy Tails
Yun, Seokhoon ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 759~766
DOI : 10.5351/KJAS.2013.26.5.759
The paper examines several classes of probability distributions with heavy tails. An (asymptotic) expression for tail probability needs to be known to understand which class a given probability distribution belongs to. It is usually not easy to get expressions for tail probabilities since most absolutely continuous probability distributions are specified by probability density functions and not by distribution functions. The paper proposes a method to obtain asymptotic expressions for tail probabilities using only probability density functions. Some examples are given to illustrate the proposed method.
Improved Generalized Method of Moment Estimators to Estimate Diffusion Models
Choi, Youngsoo ; Lee, Yoon-Dong ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 767~783
DOI : 10.5351/KJAS.2013.26.5.767
Generalized Method of Moment(GMM) is a popular estimation method to estimate model parameters in empirical financial studies. GMM is frequently applied to estimate diffusion models that are basic techniques of modern financial engineering. However, recent research showed that GMM had poor properties to estimate the parameters that pertain to the diffusion coefficient in diffusion models. This research corrects the weakness of GMM and suggests alternatives to improve the statistical properties of GMM estimators. In this study, a simulation method is adopted to compare estimation methods. Out of compared alternatives, NGMM-Y, a version of improved GMM that adopts the NLL idea of Shoji and Ozaki (1998), showed the best properties. Especially NGMM-Y estimator is superior to other versions of GMM estimators for the estimation of diffusion coefficient parameters.
A Graphical Improvement in Volatility Analysis for Financial Series
Lee, Jeong Won ; Yoon, Jae Eun ; Hwang, Sun Young ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 785~796
DOI : 10.5351/KJAS.2013.26.5.785
News Impact Curves(NIC) developed by Engle and Ng (1993) have been useful for graphically representing the volatilities arising from financial time series. Adding an improvement and refinement to the original NIC, this article proposes so called two dimensional NIC and principal component NIC. We illustrate the methodology via Kosdaq data.
New Method for Combining P-values in Meta-Analysis
Seon, Jeongyeon ; Kim, Dongjae ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 797~806
DOI : 10.5351/KJAS.2013.26.5.797
Meta-analysis is used in variety of areas to synthesize the results of previous studies. Among the methods for Meta-analysis, combining p-values is the simplest method; in addition Tippett (1931), Fisher (1932), Stuoffer at al. (1949), proposed various methods to combine p-values. We propose a new method to combine p-values based on exponential distribution. A Monte Carlo simulation study compares the power of the proposed methods with previous methods.
Exploratory Data Analysis for Korean Stock Data with Recurrence Plots
Jang, Dae-Heung ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 807~819
DOI : 10.5351/KJAS.2013.26.5.807
A recurrence plot can be used as a graphical exploratory data analysis tool before confirmatory time series analysis. With the recurrence plot, we can obtain the structural pattern of the time series and recognize the structural change points in a time series at a glance. Korean stock data shows the usefulness of the recurrence plot as a graphical exploratory data analysis tool for time series data.
Variable Selection in Clustering by Recursive Fit of Normal Distribution-based Salient Mixture Model
Kim, Seung-Gu ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 821~834
DOI : 10.5351/KJAS.2013.26.5.821
Law et al. (2004) proposed a normal distribution based salient mixture model for variable selection in clustering. However, this model has substantial problems such as the unidentifiability of components an the inaccurate selection of informative variables in the case of a small cluster size. We propose an alternative method to overcome problems and demonstrate a good performance through experiments on simulated data and real data.
A Study of Travel Time Prediction using K-Nearest Neighborhood Method
Lim, Sung-Han ; Lee, Hyang-Mi ; Park, Seong-Lyong ; Heo, Tae-Young ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 835~845
DOI : 10.5351/KJAS.2013.26.5.835
Travel-time is considered the most typical and preferred traffic information for intelligent transportation systems(ITS). This paper proposes a real-time travel-time prediction method for a national highway. In this paper, the K-nearest neighbor(KNN) method is used for travel time prediction. The KNN method (a nonparametric method) is appropriate for a real-time traffic management system because the method needs no additional assumptions or parameter calibration. The performances of various models are compared based on mean absolute percentage error(MAPE) and coefficient of variation(CV). In real application, the analysis of real traffic data collected from Korean national highways indicates that the proposed model outperforms other prediction models such as the historical average model and the Kalman filter model. It is expected to improve travel-time reliability by flexibly using travel-time from the proposed model with travel-time from the interval detectors.
A Comparison of the Effects of Optimization Learning Rates using a Modified Learning Process for Generalized Neural Network
Yoon, Yeochang ; Lee, Sungduck ;
Korean Journal of Applied Statistics, volume 26, issue 5, 2013, Pages 847~856
DOI : 10.5351/KJAS.2013.26.5.847
We propose a modified learning process for generalized neural network using a learning algorithm by Liu et al. (2001). We consider the effect of initial weights, training results and learning errors using a modified learning process. We employ an incremental training procedure where training patterns are learned systematically. Our algorithm starts with a single training pattern and a single hidden layer neuron. During the course of neural network training, we try to escape from the local minimum by using a weight scaling technique. We allow the network to grow by adding a hidden layer neuron only after several consecutive failed attempts to escape from a local minimum. Our optimization procedure tends to make the network reach the error tolerance with no or little training after the addition of a hidden layer neuron. Simulation results with suitable initial weights indicate that the present constructive algorithm can obtain neural networks very close to minimal structures and that convergence to a solution in neural network training can be guaranteed. We tested these algorithms extensively with small training sets.