Go to the main menu
Skip to content
Go to bottom
REFERENCE LINKING PLATFORM OF KOREA S&T JOURNALS
> Journal Vol & Issue
Communications for Statistical Applications and Methods
Journal Basic Information
Journal DOI :
The Korean Statistical Society
Editor in Chief :
Volume & Issues
Volume 9, Issue 3 - Dec 2002
Volume 9, Issue 2 - Aug 2002
Volume 9, Issue 1 - Apr 2002
Selecting the target year
Classification Analysis in Information Retrieval by Using Gauss Patterns
Lee, Jung-Jin ; Kim, Soo-Kwan ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 1~11
DOI : 10.5351/CKSS.2002.9.1.001
This paper discusses problems of the Poisson Mixture model which Is widely used to decide the effective words in judging relevant document. Gamma Distribution model and Gauss Patterns model as an alternative of the Poisson Mixture model are studied. Classification experiments by using TREC sub-collection, WSJ[1,2] with MGQUERY and AidSearch3.0 system are discussed.
Modified Mass-Preserving Sample Entropy
Kim, Chul-Eung ; Park, Sang-Un ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 13~19
DOI : 10.5351/CKSS.2002.9.1.013
In nonparametric entropy estimation, both mass and mean-preserving maximum entropy distribution (Theil, 1980) and the underlying distribution of the sample entropy (Vasicek, 1976), the most widely used entropy estimator, consist of nb mass-preserving densities based on disjoint Intervals of the simple averages of two adjacent order statistics. In this paper, we notice that those nonparametric density functions do not actually keep the mass-preserving constraint, and propose a modified sample entropy by considering the generalized 0-statistics (Kaigh and Driscoll, 1987) in averaging two adjacent order statistics. We consider the proposed estimator in a goodness of fit test for normality and compare its performance with that of the sample entropy.
A Bayesian Approach to Optimal Replacement Policy for a Repairable System with Warranty Period
Jung, Gi-Mun ; Han, Sung-Sil ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 21~31
DOI : 10.5351/CKSS.2002.9.1.021
This paper considers a Bayesian approach to determine an optimal replacement policy for a repairable system with warranty period. The mathematical formula of the expected cost rate per unit time is obtained for two cases : RFRW(renewing free-replacement warranty) and RPRW(renewing pro-rata warranty). When the failure time is Weibull distribution with uncertain parameters, a Bayesian approach is established to formally express and update the uncertain parameters for determining an optimal replacement policy. Some numerical examples are presented for illustrative purpose.
Optimality of Blocked Complete Diallel Crosses Using BIBD
Kim, Jin ; Bae, Jongsung ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 33~38
DOI : 10.5351/CKSS.2002.9.1.033
We prove that blocked complete dialler crosses using balance incomplete block design are A-, D-, E-optimal design. Also, we provide such blocked complete dialled crosses and their average efficiency with table.
Internet Survey System Construction and Utilization of Web log Data - APM SURVEYOR 1.5 -
Cha, Kyung-Joon ; Jung, Jae-Woo ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 39~51
DOI : 10.5351/CKSS.2002.9.1.039
In this paper, we propose a poll system on the interned. This poll system makes use of APM which are useful in web and PC using for a server. These tools are all free to obtain, so we can construct this system with the minimum computing environment and the minimum cost. We mention merits and demerits about interned survey and propose how to overcome using this system, and utilize web log data to get an additional information of panel. Finally, we suggest extensibilities of internet survey and the proposed system
Block Designs for Diallel Crosses with Different Group Characteristics
Kim, Seo-Young ; Bae, Jong-Sung ; Han, Wean-Sik ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 53~61
DOI : 10.5351/CKSS.2002.9.1.053
In this paper we proposed the method of the block designs for diallel crosses when inbred lines are divided into two groups. These block designs are derived by using kornecker product designs with two balanced incomplete block designs, and their efficiency factor evaluated. A table of diallel crosses for up to p
15 lines, p
15 lines in each set is also provided.
Estimating a Binomial Proportion with Bayes Estimated Imputed Conditional Means
Shin, Min-Woong ; Lee, Sang-Eun ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 63~73
DOI : 10.5351/CKSS.2002.9.1.063
The one of analytic imputation technique involving conditional means was mentioned by Schafer and Schenker(2000). And their derivations are based on asymptotic expansions of point estimator and their associated variance estimator, and the result of imputation can be thought of as first-order approximations to the estimators. Specially in this paper, we are presenting the method of estimating a Binomial proportion with Bayesian approach of imputed conditional means. That is, instead of using maximum likelihood(ML) estimator to estimate a Binomial proportion, in general, we use the Bayesian estimators and will show the result of estimated Imputed conditional means.
Change-point Estimation based on Log Scores
Kim, Jaehee ; Seo, Hyunjoo ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 75~86
DOI : 10.5351/CKSS.2002.9.1.075
We consider the problem of estimating the change-point in mean change model with one change-point. Gombay and Huskova(1998) derived a class of change-point estimators with the score function of rank. A change-point estimator with the log score function of rank is suggested and is shown to be involved in the class of Gombay and Huskova(1988). The simulation results show that the proposed estimator has smaller rose, larger proportion of matching the true change-point than the other estimators considered in the experiment when the change-point occurs in the middle of the sample.
A Study on the Poorly-posed Problems in the Discriminant Analysis of Growth Curve Model
Shim, Kyu-Bark ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 87~100
DOI : 10.5351/CKSS.2002.9.1.087
Poorly-posed problems in the balanced discriminant analysis was considered. We restrict consideration to the case of observations and the number of variables are the same and small. When these problems exist, we do not use a maximum likelihood estimates(MLE) to estimate covariance matrices. Instead of MLE, an alternative estimate for the covariance matrices are proposed. This alternative method make good use of two regularization parameters,
. A new test rule for the discriminant function is suggested and examined via limited hut informative simulation study. From the simulation study, it is shown that the suggested test rule gives better test result than other previously suggested method in terms of error rate criterion.
New Dispersion Function in the Rank Regression
Choi, Young-Hun ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 101~113
DOI : 10.5351/CKSS.2002.9.1.101
In this paper we introduce a new score generating (unction for the rank regression in the linear regression model. The score function compares the
'th and s\`th power of the tail probabilities of the underlying probability distribution. We show that the rank estimate asymptotically converges to a multivariate normal. further we derive the asymptotic Pitman relative efficiencies and the most efficient values of
and s under the symmetric distribution such as uniform, normal, cauchy and double exponential distributions and the asymmetric distribution such as exponential and lognormal distributions respectively
Hierarchical Bayes Analysis of Smoking and Lung Cancer Data
Oh, Man-Suk ; Park, Hyun-Jin ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 115~128
DOI : 10.5351/CKSS.2002.9.1.115
Hierarchical models are widely used for inference on correlated parameters as a compromise between underfitting and overfilling problems. In this paper, we take a Bayesian approach to analyzing hierarchical models and suggest a Markov chain Monte Carlo methods to get around computational difficulties in Bayesian analysis of the hierarchical models. We apply the method to a real data on smoking and lung cancer which are collected from cities in China.
Default Bayesian Method for Detecting the Changes in Sequences of Independent Exponential and Poisson Random Variates
Jeong, Su-Youn ; Son, Young-Sook ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 129~139
DOI : 10.5351/CKSS.2002.9.1.129
Default Bayesian method for detecting the changes in sequences of independent exponential random variates and independent Poisson random variates is considered. Noninformative priors are assumed for all the parameters in both of change models. Default Bayes factors, AIBF, MIBF, FBF, to check whether there is any change or not on each sequence and the posterior probability densities of change at each time point are derived. Theoretical results discussed in this paper are applied to some numerical data.
An Empirical Investigation on the Interactions of Foreign Investments, Stock Returns and Foreign Exchange Rates
Kim, Yoon-Tae ; Lee, Kyu-Seok ; Shin, Dong-Ho ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 141~154
DOI : 10.5351/CKSS.2002.9.1.141
Foreign investors'shares and their influences on the Korean stock market have never been larger and greater before since the market was completely open to foreign investors in 1992 Quantitatively and qualitatively as well, as a result, changes in the patterns of foreign investments have caused enormous effects on the interactions of major macroeconomic indices of the Korean economy. This paper is intended to investigate the causal relations of the four variables, foreigners'buy-sell ratios, stock returns, ￦/＄ exchange rates and
/＄ exchange rates, over the two time periods of the pre-IMF (1996.1.1-1997.8.15) and the post-IMF (1997.8.16-2000.6.15) based on the daily data of the variables. Granger Causality Test, Forecast Error Variance Decomposition(FEVD) using VAR model and Impulse Response Function were implemented for the empirical analysis.
Bayesian Analysis for Neural Network Models
Chung, Younshik ; Jung, Jinhyouk ; Kim, Chansoo ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 155~166
DOI : 10.5351/CKSS.2002.9.1.155
Neural networks have been studied as a popular tool for classification and they are very flexible. Also, they are used for many applications of pattern classification and pattern recognition. This paper focuses on Bayesian approach to feed-forward neural networks with single hidden layer of units with logistic activation. In this model, we are interested in deciding the number of nodes of neural network model with p input units, one hidden layer with m hidden nodes and one output unit in Bayesian setup for fixed m. Here, we use the latent variable into the prior of the coefficient regression, and we introduce the 'sequential step' which is based on the idea of the data augmentation by Tanner and Wong(1787). The MCMC method(Gibbs sampler and Metropolish algorithm) can be used to overcome the complicated Bayesian computation. Finally, a proposed method is applied to a simulated data.
On the Design of Statistical Software in the Network Environment
Han, Beom-Soo ; Ahn, Jeong-Yong ; Han, Kyung-Soo ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 167~174
DOI : 10.5351/CKSS.2002.9.1.167
Computer network provides a powerful infrastructure for information sharing and the development of the statistical software with new concepts. In this paper, we discuss the design concepts of the statistical software in the network environment.
Efficient Estimation of Population Mean Using Centered Modified Systematic Sampling and Interpolation
Kim, Hyuk-Joo ; Choi, Byoung-Chul ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 175~185
DOI : 10.5351/CKSS.2002.9.1.175
A method is proposed for efficiently estimating the mean of a population which has a linear trend. The proposed estimator is based on the centered modified systematic sampling method and the concept of interpolation. Using the expected mean square error criterion, it is shown that the proposed method is more efficient than conventional methods in most real cases.
The Limit Distribution and Power of a Test for Bivariate Normality
Kim, Namhyun ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 187~196
DOI : 10.5351/CKSS.2002.9.1.187
Testing for normality has always been a center of practical and theoretical interest in statistical research. In this paper a test statistic for bivariate normality is proposed. The underlying idea is to investigate all the possible linear combinations that reduce to the standard normal distribution under the null hypothesis and compare the order statistics of them with the theoretical normal quantiles. The suggested statistic is invariant with respect to nonsingular matrix multiplication and vector addition. We show that the limit distribution of an approximation to the suggested statistic is represented as the supremum over an index set of the integral of a suitable Gaussian Process. We also simulate the null distribution of the statistic and give some critical values of the distribution and power results.
Influential Points in GLMs via Backwards Stepping
Jeong, Kwang-Mo ; Oh, Hae-Young ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 197~212
DOI : 10.5351/CKSS.2002.9.1.197
When assessing goodness-of-fit of a model, a small subset of deviating observations can give rise to a significant lack of fit. It is therefore important to identify such observations and to assess their effects on various aspects of analysis. A Cook's distance measure is usually used to detect influential observation. But it sometimes is not fully effective in identifying truly influential set of observations because there may exist masking or swamping effects. In this paper we confine our attention to influential subset In GLMs such as logistic regression models and loglinear models. We modify a backwards stepping algorithm, which was originally suggested for detecting outlying cells in contingency tables, to detect influential observations in GLMs. The algorithm consists of two steps, the identification step and the testing step. In identification step we Identify influential observations based on influencial measures such as Cook's distances. On the other hand in testing step we test the subset of identified observations to be significant or not Finally we explain the proposed method through two types of dataset related to logistic regression model and loglinear model, respectively.
Estimation of Hard-to-Measure Measurements in Anthropometric Surveys
Choi, Jong-Hoo ; Kim, Ryu-Jin ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 213~220
DOI : 10.5351/CKSS.2002.9.1.213
Anthropometric survey is important as a basis for human engineering fields. According to our experiences, there are difficulties in obtaining the measurements of some body parts because respondents are reluctant to expose. In order to overcome these difficulties, we propose a method for estimating such hard-to-measure measurements by using easy-to-measure measurements those are closely related to them. Multiple Regression Model, Feedforward Neural Network(FNN) Model and Projection Pursuit Regression(PPR) Model will be used as analytical tools for this purpose. The method we propose will be illustrated with real data from the 1992 Korea national anthropometric survey.
A Study on the Role of Pivots in Bayesian Statistics
Hwang, Hyungtae ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 221~227
DOI : 10.5351/CKSS.2002.9.1.221
The concept of pivot has been widely used in various classical inferences. In this paper, it is proved by use of pivotal quantities that the Bayesian inferences can be arrived at the same results of classical inferences for the location-scale parameters models under the assumption of non-informative prior distributions. Some theorems are proposed in which the posterior distribution and the sampling distribution of a pivotal quantity coincide. The theorems are applied illustratively to some statistical models.
Comparison of Powers in Goodness of Fit Test of Quadratic Measurement Error Model
Moon, Myung-Sang ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 229~240
DOI : 10.5351/CKSS.2002.9.1.229
Whether to use linear or quadratic model in the analysis of regression data is one of the important problems in classical regression model and measurement error model (MEM). In MEM, four goodness of fit test statistics are available In solving that problem. Two are from the derivation of estimators of quadratic MEM, and one is from that of the general
-order polynomial MEM. The fourth one is derived as a variation of goodness of fit test statistic used in linear MEM. The purpose of this paper is to find the most powerful test statistic among them through the small-scale simulation.
A Note on Central Limit Theorem for Deconvolution Wavelet Density Estimators
Lee, Sungho ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 241~248
DOI : 10.5351/CKSS.2002.9.1.241
The problem of wavelet density estimation based on Shannon's wavelets is studied when the sample observations are contaminated with random noise. In this paper we will discuss the asymptotic normality for deconvolving wavelet density estimator of the unknown density f(x) when courier transform of random noise has polynomial descent.
Estimation of the Mean and Variance for Normal Distributions whose Both Sides are Truncated
Hong, Chong-Sun ; Choi, Yun-Young ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 249~259
DOI : 10.5351/CKSS.2002.9.1.249
In order to estimate the mean and variance for a Normal distribution which is truncated at both right and left sides, maximum likelihood estimators based on the entire sample from the original distribution are compared with the sample mean and variance of the censored sample which is the data remaining after truncation using simulation. We found that, surprisingly, the mean squared error of the mean based on the censored data Is smaller than that of the full sample estimators.
Confidence Intervals for the Median Survival Time under Proportional Censorship
Jeong, Seong-Hwa ; Cho, Kil-Ho ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 261~270
DOI : 10.5351/CKSS.2002.9.1.261
In this paper, we demonstrate the more accurate confidence intervals for median survival time under the simple proportional hazard model of Koziol and Green (1976) via the Edgeworth expansion for the distribution of the studentized ACL estimator derived in Jeong (2000). The numerical results show that the intervals, so-called test-based and reflect intervals (Slud et al., 1984), outperform normal approximating method in the small sample sizes and/or heavy censoring.
Development of Scoring Model on Customer Attrition Probability by Using Data Mining Techniques
Han, Sang-Tae ; Lee, Seong-Keon ; Kang, Hyun-Cheol ; Ryu, Dong-Kyun ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 271~280
DOI : 10.5351/CKSS.2002.9.1.271
Recently, many companies have applied data mining techniques to promote competitive power in the field of their business market. In this study, we address how data mining, that is a technique to enable to discover knowledge from a deluge of data, Is used in an executed project in order to support decision making of an enterprise. Also, we develope scoring model on customer attrition probability for automobile-insurance company using data mining techniques. The development of scoring model in domestic insurance is given as an example concretely
Bootstrapping Logit Model
Kim, Dae-hak ; Jeong, Hyeong-Chul ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 281~289
DOI : 10.5351/CKSS.2002.9.1.281
In this paper, we considered an application of the bootstrap method for logit model. Estimation of type I error probability, the bootstrap p-values and bootstrap confidence intervals of parameter were proposed. Small sample Monte Carlo simulation were conducted in order to compare proposed method with existing normal theory based asymptotic method.
A Hybrid Algorithm for Identifying Multiple Outlers in Linear Regression
Kim, Bu-yong ; Kim, Hee-young ;
Communications for Statistical Applications and Methods, volume 9, issue 1, 2002, Pages 291~304
DOI : 10.5351/CKSS.2002.9.1.291
This article is concerned with an effective algorithm for the identification of multiple outliers in linear regression. It proposes a hybrid algorithm which employs the least median of squares estimator, instead of the least squares estimator, to construct an Initial clean subset in the stepwise forward search scheme. The performance of the proposed algorithm is evaluated and compared with the existing competitor via an extensive Monte Carlo simulation. The algorithm appears to be superior to the competitor for the most of scenarios explored in the simulation study. Particularly it copes with the masking problem quite well. In addition, the orthogonal decomposition and Its updating techniques are considered to improve the computational efficiency and numerical stability of the algorithm.