• The Korean Journal of Applied Statistics
• /
• v.7 no.2
• /
• pp.9-19
• /
• 1994

Beta-binomial model, which is reparametrized in terms of the mean probability $\mu$ of a positive deagnosis and the $\kappa$ of agreement, is widely used in psychology. When $\mu$ is close to 0, inference about $\kappa$ become difficult because likelihood function becomes constant. We consider Bayesian approach in this case. To apply Bayesian analysis, Gibbs sampler is used to overcome difficulties in integration. Marginal posterior density functions are estimated and Bayesian estimates are derived by using Gibbs sampler and compare the results with the one obtained by using numerical integration.

• The Korean Journal of Applied Statistics
• /
• v.9 no.1
• /
• pp.111-124
• /
• 1996

In this paper, we consider the first-order integer valued autoregressive(INAR(1)) model where correlation structure is similar to that of the continuous valued AR(1) process. Several methods for estimating the parameters of the INAR(1) process with negative binomial marginal are discussed. We derive asymptotic distributions of these estimators. The results of a simulation study for these estimators methods show that the estimator which we present in this paper is better than the estimator which Klimko and Nelson(1978) presented. As an application we considered the estimator of M/M/1 queue length.

• Bulletin of the Korean Mathematical Society
• /
• v.60 no.4
• /
• pp.1017-1024
• /
• 2023

A celebrated result in the study of integer partitions is the identity due to Lehmer whereby the number of partitions of n with an even number of even parts minus the number of partitions of n with an odd number of even parts equals the number of partitions of n into distinct odd parts. Inspired by Lehmer's identity, we prove explicit formulas for evaluating generating functions for sequences that enumerate integer partitions of fixed width with an even/odd number of even parts. We introduce a technique for decomposing the even entries of a partition in such a way so as to evaluate, using a finite sum over q-binomial coefficients, the generating function for the sequence of partitions with an even number of even parts of fixed, odd width, and similarly for the other families of fixed-width partitions that we introduce.

• Environmental and Resource Economics Review
• /
• v.27 no.3
• /
• pp.545-567
• /
• 2018

This study applied and compared Poisson model, negative binomial model, zero inflated Poisson model, and zero inflated negative binomial model to estimate determinants of employed labour quantity. To estimate each of models, this study used fisheries census data which were obtained at microdata integrated service running by Statistics Korea. The study selected zero inflated negative binomial model according to the Vuong test and Likelihood-ratio test. In addition, the study estimated fishing village's practical changes on employed labour quantity as analyzing changes from 2010 to 2015. The results showed that the household with fishing vessels and high selling price had a significant effect on decrease of the labour quantities. Meanwhile, the longer work experience of the household, the more significant the increase in the labour quantities. In conclusion, this study presented that capitalized fishing household and the acceleration of aging had a significant impact on the change in the labour quantities.

• The Korean Journal of Applied Statistics
• /
• v.34 no.5
• /
• pp.735-744
• /
• 2021

Maintaining the lifetime of a product is one of the objectives of quality control. In real processes, most samples are constructed with censored data because, in many situations, we cannot measure the lifetime of all samples due to time or cost problems. In this paper, we propose two cumulative sum (CUSUM) control charting procedures to monitor the mean of type I right-censored lognormal lifetime data. One of them is based on the likelihood ratio, and the other is based on the binomial distribution. Through simulations, we evaluate the performance of the two proposed procedures by comparing the average run length (ARL). The overall performance of the likelihood ratio CUSUM chart is better, especially this chart performs better when the censoring rate is low and the shape parameter value is small. Conversely, the binomial CUSUM chart is shown to perform better when the censoring rate is high, the shape parameter value is large, and the change in the mean is small.

• The Korean Journal of Applied Statistics
• /
• v.34 no.5
• /
• pp.711-721
• /
• 2021

In this paper, we propose the method of adjusting thresholds using impurity indices in classification analysis on imbalanced data. Suppose the minority category is Positive and the majority category is Negative for the imbalanced binomial data. When categories are determined based on the commonly used 0.5 basis, the specificity tends to be high in unbalanced data while the sensitivity is relatively low. Increasing sensitivity is important when proper classification of objects in minority categories is relatively important. We explore how to increase sensitivity through adjusting thresholds. Existing studies have adjusted thresholds based on measures such as G-Mean and F1-score, but in this paper, we propose a method to select optimal thresholds using the chi-square statistic of CHAID, the Gini index of CART, and the entropy of C4.5. We also introduce how to get a possible unique value when multiple optimal thresholds are obtained. Empirical analysis shows what improvements have been made compared to the results based on 0.5 through classification performance metrics.

• Journal of Korean Society for Quality Management
• /
• v.25 no.3
• /
• pp.62-73
• /
• 1997

Previous VSI control chart works have been done on quality variable whose distribution is normal. But there are many situations in which hte assumption of not a, pp.opriate. Also, in many industrial processes, the interest is to monitor the number of defectives. In this paper, we will take the existing properties of VSI control chart developed for the normal distribution and a, pp.y them to the np-chart based on the discrete binomial distribution. We will consider the CUSUM chart for the number of defectives. Here, the interesting object is to compute the VSI ATS for CUSUM control chart using Markov chain a, pp.oach and to compare FSI ATS and VSI ATS.

• The Korean Journal of Applied Statistics
• /
• v.23 no.5
• /
• pp.883-890
• /
• 2010

Since Fisher's exact test is conducted conditional on the observed value of the margin, there are two kinds of the exact power, the conditional and the unconditional exact power. The conditional exact power is computed at a given value of the margin whereas the unconditional exact power is calculated by incorporating the uncertainty of the margin. Although the sample size is determined based on the unconditional exact power, the actual power which Fisher's exact test has is the conditional power after the experiment is finished. This paper investigates differences between the conditional and unconditional exact power Fisher's exact test. We conclude that such discrepancy is a disadvantage of Fisher's exact test.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.11a
• /
• pp.181-186
• /
• 2005

Many data sets obtained from surveys or medical trials often include missing observations. When these data sets are analyzed, it is general to use only complete cases. However, it is possible to have big biases or involve inefficiency. In this paper, we consider a method for estimating parameters in logistic linear models involving non-ignorable missing data mechanism. A binomial response and normal exploratory model for the missing data are used. We fit the model using the EM algorithm. The E-step is derived by Metropolis-hastings algorithm to generate a sample for missing data and Monte-carlo technique, and the M-step is by Newton-Raphson to maximize likelihood function. Asymptotic variances of the MLE's are derived and the standard error and estimates of parameters are compared.

• Communications for Statistical Applications and Methods
• /
• v.2 no.1
• /
• pp.112-121
• /
• 1995

Let $p_1$ and $p_2$ be the proportions of two populations. To test the hypothesis $H_0 : p_1 = p_2$, we usually use the $x^2$ statistic, the large sample binomial statistic Z, and the Generalized Likelihood Ratio statistic-2log $\lambda$developed based on different mathematical rationale, respectively. Since testing the above hypothesis is equivalent to testing whether two populations follow the common Bernoulli distribution, one may also test the hypothesis by comparing 1 with the ratio of each density estimate and the hypothesized common density estimate, called comparison density, which was devised by Parzen(1988). We show that the above traditional test statistics ate actually estimating the measure of distance between the true densities and the common density under $H_0$ by representing them with the comparison density.