• Journal of the Korean Data and Information Science Society
• /
• v.14 no.2
• /
• pp.177-186
• /
• 2003

The Zero-Inflated Poisson regression is a model for count data with exess zeros. When the correlated response variables are intrested, we have to extend the univariate zero-inflated regression model to multivariate model. In this paper, we study and simulate the multivariate zero-inflated regression model. A real example was applied to this model. Regression parameters are estimated by using MLE's. We also compare the fitness of multivariate zero-inflated Poisson regression model with the decision tree model.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.34 no.1
• /
• pp.42-51
• /
• 2011

Poisson model and Gamma-Poisson model are popularly used to analyze statistical behavior from defective data. The methods are based on binary criteria, that is, good or failure. However, manufacturing industries prefer polytomous criteria for classifying manufactured products due to flexibility of marketing. In this paper, I introduce two multivariate Gamma-Poisson(MGP) models and estimation methods of the parameters in the models, which are able to handle polytomous data. The models and estimators are verified on defective pixels of LCD manufacturing. Experimental results show that both the independent MGP model and the multinomial MGP model have excellent performance in terms of mean absolute deviation and the choice of method depends on the purpose of use.

• Korean Management Science Review
• /
• v.32 no.2
• /
• pp.69-78
• /
• 2015

This paper considers a Bayesian Poisson model for multivariate count data using multiplicative rates. More specifically we compose the parameter for overall arrival rates by the product of two parameters, a common effect and an individual effect. The common effect is composed of autoregressive evolution of the parameter, which allows for analysis on seasonal effects on all multivariate time series. In addition, analysis on individual effects allows the researcher to differentiate the time series by whatevercharacterization of their choice. This type of model allows the researcher to specifically analyze two different forms of effects separately and produce a more robust result. We illustrate a simple MCMC generation combined with a Gibbs sampler step in estimating the posterior joint distribution of all parameters in the model. On the whole, the model presented in this study is an intuitive model which may handle complicated problems, and we highlight the properties and possible applications of the model with an example, analyzing real time series data involving customer arrivals to a large retail store.

• Communications for Statistical Applications and Methods
• /
• v.11 no.2
• /
• pp.381-397
• /
• 2004

The marginal likelihood has become an important tool for model selection in Bayesian analysis because it can be used to rank the models. We discuss the marginal likelihood for Poisson regression models that are potentially useful in small area estimation. Computation in these models is intensive and it requires an implementation of Markov chain Monte Carlo (MCMC) methods. Using importance sampling and multivariate density estimation, we demonstrate a computation of the marginal likelihood through an output analysis from an MCMC sampler.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.2
• /
• pp.187-194
• /
• 2014

In this paper, we analyze the outage probability of macro diversity combining in cellular networks in consideration of aggregate interference from other mobile stations (MSs). Different from existing works analyzing the outage probability of macro diversity combining, we focus on a diversity gain attained by selecting a base station (BS) subject to relatively low aggregate interference. In our model, MSs are randomly located according to a Poisson point process. The outage probability is analyzed by approximating the multivariate distribution of aggregate interferences on multiple BSs by a multivariate lognormal distribution.

• The Korean Journal of Applied Statistics
• /
• v.22 no.4
• /
• pp.687-695
• /
• 2009

The Bayesian multiple change-point estimation has been applied to the daily means of ozone and PM10 data in Seoul for the period 1999. We focus on the detection of multiple change-points in the ozone and PM10 bivariate vectors by evaluating the posterior probabilities and Bayesian information criterion(BIC) using the stochastic approximation Monte Carlo(SAMC) algorithm. The result gives 5 change-points of mean vectors of ozone and PM10, which are related with the seasonal characteristics.

• Journal of Preventive Medicine and Public Health
• /
• v.52 no.3
• /
• pp.147-153
• /
• 2019

Objectives: The purpose of this study was to evaluate the association between body mass index (BMI) and severe lower urinary tract symptoms (LUTS) in Korean males. Methods: This study was conducted on males aged ${\geq}50years$ who participated in the 2011 Korean Community Health Survey. LUTS severity was assessed using the Korean version of the International Prostate Symptom Score (IPSS) questionnaire, and was dichotomized as severe (IPSS >19) and non-severe ($IPSS{\leq}19$). BMI was divided into 6 categories: <18.5, 18.5-22.9, 23.0-24.9, 25.0-27.4, 27.5-29.9, and ${\geq}30.0kg/m^2$. To evaluate the relationship between BMI and LUTS, a survey-weighted multivariate Poisson regression analysis was performed to estimate prevalence rate ratios (PRRs). Age, smoking status, alcohol intake, physical activity, educational level, household income, and comorbidities were adjusted for in the multivariate model. Results: A U-shaped relationship was detected between BMI and severe LUTS. Compared with a BMI of $23.0-24.9kg/m^2$, the PRR for a BMI < $18.5kg/m^2$ was 1.65 (95% confidence interval [CI], 1.35 to 2.02), that for a BMI of $18.5-22.9kg/m^2$ was 1.25 (95% CI, 1.09 to 1.44), that for a BMI of $25.0-27.4kg/m^2$ was 1.20 (95% CI, 1.00 to 1.45), that for a BMI of $27.5-29.9kg/m^2$ was 1.11 (95% CI, 0.83 to 1.47), and that for a BMI ${\geq}30.0kg/m^2$ was 1.85 (95% CI, 1.18 to 2.88). Conclusions: This study showed that both high and low BMI were associated with severe LUTS.

• Journal of Preventive Medicine and Public Health
• /
• v.46 no.2
• /
• pp.89-95
• /
• 2013

Objectives: Little is yet known about the determinants of bone mineral density (BMD) in young adults. Thus, in this study, we aimed to determine the factors that have an impact on BMD in young men. Methods: Questionnaires were sent out to 111 male medical students. Information on age, socio-economic status, medical history, lifestyle, physical activity during adolescence, school club participation, current physical activity, and dietary intake were collected by the survey. Height, weight, percent body fat and muscle mass were estimated by bioelectrical impedance, and BMD was obtained using calcaneal quantitative ultrasound. Using the Poisson regression model, prevalence ratios (PRs) were used to estimate the degree of association between risk factors and osteopenia. Results: The height and current physical activity showed a correlation to the Osteoporosis Index. Among the categorized variables, past physical activity during adolescence (p= 0.002) showed a positive effect on the bone mineral content. In the multivariate model, past physical activity (${\geq}1$ time/wk) had a protective effect on osteopenia (PR, 0.37; 95% confidence interval [CI], 0.18 to 0.75) and present physical activity (1000 metabolic equivalent of task-min/wk) decreased the risk of osteopenia (PR, 0.64; 95% CI, 0.44 to 0.91). Conclusions: Past physical activity during adolescence is as important as physical activity in the present for BMD in young men.

• Nutrition Research and Practice
• /
• v.9 no.2
• /
• pp.207-212
• /
• 2015

BACKGROUND/OBJECTIVES: The Recaller app was developed to help individuals record their food intakes. This pilot study evaluated the usability of this new food picture application (app), which operates on a smartphone with an embedded camera and Internet capability. SUBJECTS/METHODS: Adults aged 19 to 28 years (23 males and 22 females) were assigned to use the Recaller app on six designated, nonconsecutive days in order to capture an image of each meal and snack before and after eating. The images were automatically time-stamped and uploaded by the app to the Recaller website. A trained nutritionist administered a 24-hour dietary recall interview 1 day after food images were taken. Participants' opinions of the Recaller app and its usability were determined by a follow-up survey. As an evaluation indicator of usability, the number of images taken was analyzed and multivariate Poisson regression used to model the factors determining the number of images sent. RESULTS: A total of 3,315 food images were uploaded throughout the study period. The median number of images taken per day was nine for males and 13 for females. The survey showed that the Recaller app was easy to use, and 50% of the participants would consider using the app daily. Predictors of a higher number of images were as follows: greater interval (hours) between the first and last food images sent, weekend, and female. CONCLUSIONS: The results of this pilot study provide valuable information for understanding the usability of the Recaller smartphone food picture app as well as other similarly designed apps. This study provides a model for assisting nutrition educators in their collection of food intake information by using tools available on smartphones. This innovative approach has the potential to improve recall of foods eaten and monitoring of dietary intake in nutritional studies.