• Journal of the Korean Statistical Society
• /
• v.30 no.1
• /
• pp.1-20
• /
• 2001

We present how bootstrap methods can be used to conduct inference on the rates of Poisson distributions when only the grouped data are available. A theoretical justification for the validity of bootstrap is given with an illustration of proposed method using a data set obtained fro ma pathology laboratory test. Traditional asymptotic methods are compared with bootstrap methods in computing the estimated standard errors and achieved significance levels for one sample and two sample tests. Bootstrap methods are shown to possess a nice property that he small sample distribution of the relevant statistics can be readily obtained from the bootstrap copies.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1996.10a
• /
• pp.31-34
• /
• 1996

The extensive researches have compared the performance of neural networks(NN) with those of various statistical techniques for the classification problem. The empirical results of these comparative studies have indicated that the neural networks often outperform the traditional statistical techniques. Moreover, there are some efforts that try to combine various classification methods, especially multivariate discriminant analysis with neural networks. While these efforts improve the performance, there exists a problem violating robust assumptions of multivariate discriminant analysis that are multivariate normality of the independent variables and equality of variance-covariance matrices in each of the groups. On the contrary, cluster analysis alleviates this assumption like neural networks. We propose a new approach to classification problems by combining the cluster analysis with neural networks. The resulting predictions of the composite model are more accurate than each individual technique.

• Asia pacific journal of information systems
• /
• v.15 no.2
• /
• pp.51-63
• /
• 2005

There has been substantial research into the bankruptcy prediction. Many researchers used the statistical method in the problem until the early 1980s. Since the late 1980s, Artificial Intelligence(AI) has been employed in bankruptcy prediction. And many studies have shown that artificial neural network(ANN) achieved better performance than traditional statistical methods. However, despite ANN's superior performance, it has some problems such as overfitting and poor explanatory power. To overcome these limitations, this paper suggests a relatively new machine learning technique, support vector machine(SVM), to bankruptcy prediction. SVM is simple enough to be analyzed mathematically, and leads to high performances in practical applications. The objective of this paper is to examine the feasibility of SVM in bankruptcy prediction by comparing it with ANN, logistic regression, and multivariate discriminant analysis. The experimental results show that SVM provides a promising alternative to bankruptcy prediction.

• Communications for Statistical Applications and Methods
• /
• v.16 no.4
• /
• pp.713-722
• /
• 2009

The estimation of odds ratio and corresponding confidence intervals for case-control data have been done by traditional generalized linear models which assumed that the logarithm of odds ratio is linearly related to risk factors. We adapt a lower-dimensional approximation of Gu and Kim (2002) to provide a faster computation in nonparametric method for the estimation of odds ratio by allowing flexibility of the estimating function and its Bayesian confidence interval under the Bayes model for the lower-dimensional approximations. Simulation studies showed that taking larger samples with the lower-dimensional approximations help to improve the smoothing spline estimates of odds ratio in this settings. The proposed method can be used to analyze case-control data in medical studies.

• Journal of Korean Institute of Industrial Engineers
• /
• v.41 no.6
• /
• pp.521-529
• /
• 2015

Statistical process control techniques have been greatly implemented in industries for improving product quality and saving production costs. As a primary tool among these techniques, control charts are widely used to detect the occurrence of assignable causes. In most works on the control charts it considered the problem of monitoring the mean and variance, and the quality characteristic of interest is normally distributed. In some situations monitoring of the minimum and maximum values is more important and the quality characteristic of interest is the Weibull distribution rather than a normal distribution. In this paper, we consider the statistical design of minimum and maximum control charts when the distribution of the quality characteristic of interest is Weibull. The proposed minimum and maximum control charts are applied to the wind data. The results of the application show that the proposed method is more effective than traditional methods.

• Korean Journal of Oriental Medicine
• /
• v.13 no.3
• /
• pp.53-67
• /
• 2007

Seoul Bochun Medical Clinic Examined Patient Table' is a statistical data that Kim Young-Boon [1882-1974] put together himself. A total of 59353 cases from 1914 to 1935 were classified according to year and month of examination, and gender. Of the cases 26497 (${\fallingdotseq}$45%) were women and the remaining 32856 (${\fallingdotseq}$55%) cases were men. The most frequent disease examined was cold, followed by diarrhea and stomachache. Of the 639 different kinds of diseases ever examined, 71 had occurred on more than 100 occasions. The percentage that these 71 diseases occupies is an overwhelming 88.75%.

• The Korean Journal of Applied Statistics
• /
• v.16 no.2
• /
• pp.441-454
• /
• 2003

We have shown how e-Learning can be implemented for statistics and lab course with minimum cost of operation using APM. We have illustrated how traditional class educational component can be implemented under e-Learning environment, especially statistical package labs with ASP and web-folder.

• Journal of the Korean Statistical Society
• /
• v.26 no.4
• /
• pp.477-491
• /
• 1997

A new predictive classification rule for assigning future cases into one of two multivariate normal population (with unknown normal mixture model) is considered. The development involves calculation of posterior probability of each possible normal-mixture model via a default Bayesian test criterion, called intrinsic Bayes factor, and suggests predictive distribution for future cases to be classified that accounts for model uncertainty by weighting the effect of each model by its posterior probabiliy. In this paper, our interest is focused on constructing the classification rule that takes care of uncertainty about the types of covariance matrices (homogeneity/heterogeneity) involved in the model. For the constructed rule, a Monte Carlo simulation study demonstrates routine application and notes benefits over traditional predictive calssification rule by Geisser (1982).

• The Korean Journal of Applied Statistics
• /
• v.21 no.6
• /
• pp.1007-1014
• /
• 2008

Usability evaluation(UE) experiments are conducted to provide UE practitioners with guidelines for better outcomes. In UE research, significant quantities of empirical results have been accumulated in the past decades. While those results have been anticipated to integrate for producing generalized guidelines, traditional meta-analysis has limitations to combine UE empirical results that often show considerable heterogeneity. In this study, a new data mining method called weighted bi-clustering(WBC) was proposed to partition heterogeneous studies into homogeneous subsets. We applied the WBC to UE empirical results and identified two homogeneous subsets, each of which can be meta-analyzed. In addition, interactions between experimental conditions and UE methods were hypothesized based on the resulting partition and some interactions were confirmed via statistical tests.

• Communications for Statistical Applications and Methods
• /
• v.10 no.1
• /
• pp.211-224
• /
• 2003

Good's lambda test, a permutation test used to detect the changes of microorganism composition under two pathological conditions, has been quite popular for studying the micro-flora responsible for periodontal disease. A vast number of different micro-flora in the mouth renders the traditional chi-square test inapplicable. The main purpose of this paper is to evaluate the power of this test so that the sample size can be determined at the design stage. The robustness of this test and its comparison to two other intuitive tests are also presented. It is found that a permutation test based on likelihood ratio is more powerful than the lambda test in our simulated cases.