• The Korean Journal of Applied Statistics
• /
• v.29 no.4
• /
• pp.627-641
• /
• 2016

In this paper, we consider a seemingly unrelated regression (SUR) model and propose a nonparametric Bayesian approach to SUR with a Dirichlet process mixture of normals for modeling an unknown error distribution. Posterior distributions are derived based on the proposed model, and the posterior inference is performed via Markov chain Monte Carlo methods based on the collapsed Gibbs sampler of a Dirichlet process mixture model. We present a simulation study to assess the performance of the model. We also apply the model to precipitation data over South Korea.

• The Korean Journal of Applied Statistics
• /
• v.33 no.4
• /
• pp.365-377
• /
• 2020

We discuss latent factor regression when constructing a common structure inherent among explanatory variables to solve multicollinearity and use them as regressors to construct a linear model of a response variable. Bayesian estimation with LASSO prior of a large penalty parameter to construct a significant factor loading matrix of intrinsic interests among infinite latent structures. The estimated factor loading matrix with estimated other parameters can be inversely transformed into linear parameters of each explanatory variable and used as prediction models for new observations. We apply the proposed method to Product Service Management data of HBAT and observe that the proposed method constructs the same factors of general common factor analysis for the fixed number of factors. The calculated MSE of predicted values of Bayesian latent factor regression model is also smaller than the common factor regression model.

• The Korean Journal of Applied Statistics
• /
• v.19 no.3
• /
• pp.505-519
• /
• 2006

We consider zero-inflated count data, which is discrete count data but has too many zeroes compared to the Poisson distribution. Zero-inflated data can be found in various areas. Despite its increasing importance in practice, appropriate statistical inference on zero-inflated data is limited. Classical inference based on a large number theory does not fit unless the sample size is very large. And regular Poisson model shows lack of St due to many zeroes. To handle the difficulties, a mixture of distributions are considered for the zero-inflated data. Specifically, a mixture of a point mass at zero and a Poisson distribution is employed for the data. In addition, when there exist meaningful covariates selected to the response variable, loglinear link is used between the mean of the response and the covariates in the Poisson distribution part. We propose a Bayesian inference for the zero-inflated Poisson regression model by using a Markov Chain Monte Carlo method. We applied the proposed method to a Korean oral hygienic data and compared the inference results with other models. We found that the proposed method is superior in that it gives small parameter estimation error and more accurate predictions.

• Journal of the Korean Data and Information Science Society
• /
• v.22 no.5
• /
• pp.911-919
• /
• 2011

In recent medical analysis, it becomes more important to looking for risk factors related to mental illness. If we find and identify their relevant characteristics of the risk factors, the disease can be prevented in advance. Moreover, the study can be helpful to medical development. These kinds of studies of risk factors for mental illness have mainly been discussed by using the logistic regression model. However in this paper, data mining techniques such as CART, C5.0, logistic, neural networks and Bayesian network were used to search for the risk factors. The Bayesian network of the above data mining methods was selected as most optimal model by applying delirium data. Then, Bayesian network analysis was used to find risk factors and the relationship between the risk factors are identified through a radial graph.

• Proceedings of the Korea Inteligent Information System Society Conference
• /
• 2006.06a
• /
• pp.195-202
• /
• 2006

본 연구에서는 보다 효과적인 기업부도예측을 위하여, 동계적 방법과 인공지능 방법을 결합한 통합모형을 제시하였다. 이를 위하여 통계적인 모형 중에서 가장 널리 활용되고 있는 다변량 판별분석, 로지스틱 회귀분석과 인공 지능적인 방법으로서 최근 널리 사용되고 있는 인공신경망, 규칙유도기법, 베이지안 망의 5가지 방법론을 통합한 Voting with Performance & Weights from ANN(WP-ANN) 통합모형을 제시하였다. 실험결과, 본 연구에서 제안한 WP-ANN 통합모형은 다변량 판별분석, 로지스탁 회귀분석, 인공신경망, 규칙유도기법, 베이지안 망 등의 단일모형과 비교한 결과 가장 예측정확성이 유수한 것으로 나타났다. 따라서 본 연구를 통해 기업부도예측에 있어서 WP-ANN 통합모형이 기존의 모형들에 비해 우수한 예측정확성을 나타냄을 알 수 있었다.

• The Korean Journal of Applied Statistics
• /
• v.27 no.6
• /
• pp.1039-1047
• /
• 2014

An autoregressive model with normal errors is a natural model that attempts to fit time series data. More flexible models that include normal distribution as a special case are necessary because they can cover normality to non-normality models. The skewed exponential power distribution is a possible candidate for autoregressive models errors that may have tails lighter(platykurtic) or heavier(leptokurtic) than normal and skewness; in addition, the use of skewed exponential power distribution can reduce the influence of outliers and consequently increases the robustness of the analysis. We use SIR algorithm and grid method for an efficient Bayesian estimation.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2019.05a
• /
• pp.22-22
• /
• 2019

최근 전 세계적으로 급증하는 기후변화의 영향으로 인해 강우량 증가에 따른 이상홍수 발생 및 댐 여유고 부족 등 다양한 위험인자가 노출되고 있다. 이러한 예상치 못한 이상홍수는 실제 거주하고 있는 사람들을 위협할 수 있으며, 하천 범람으로 인해 2차 3차 피해가 일어날 가능성이 존재하고 있다. 이에 다양한 자연재해로부터 인명 및 재산 피해를 방지 및 저감하기 위한 목적으로 다양한 수공구조물이 존재하며, 수자원 관리계획 수립의 목적에 따라 다양한 강수량이 활용되고 있다. 특히, 지구온난화에 따른 기후변화 영향을 고려한 연최대 강수량 및 확률강수량 산정이 필요한 시점이며, 온도변화에 따른 증기압 계산식인 Clausius-Clapeyron 관계에 따르면 대기 온도가 $1^{\circ}C$ 상승할 때 대기수분량이 6~7% 증가하여 평균 온도상승에 따라 극치강수량 발생 잠재력이 향상 될 것으로 전망되고 있다. 본 연구에서는 온도상승에 따른 극치강수량의 변화를 베이지안 다중분위회귀분석모형을 통해 산정하여 CORDEX 온도자료 기반의 미래 극치강수량을 전망하였다. 본 연구결과 100년 이상 빈도의 강수량은 온도상승에 따라 급격히 증가하는 추세를 확인하였으며, 2100년까지 온도상승을 고려한 최대 극치강수량은 1500mm를 넘을 가능성을 확인하였다.

• The Korean Journal of Applied Statistics
• /
• v.31 no.2
• /
• pp.287-301
• /
• 2018

It is common to encounter count data with excess zeros in various research fields such as the social sciences, natural sciences, medical science or engineering. Such count data have been explained mainly by zero-inflated Poisson model and extended models. Zero-inflated count data are also often correlated or clustered, in which random effects should be taken into account in the model. Frequentist approaches have been commonly used to fit such data. However, a Bayesian approach has advantages of prior information, avoidance of asymptotic approximations and practical estimation of the functions of parameters. We consider a Bayesian zero-inflated Poisson regression model with random effects for correlated zero-inflated count data. We conducted simulation studies to check the performance of the proposed model. We also applied the proposed model to smoking behavior data from the Regional Health Survey (2015) of the Korea Centers for disease control and prevention.

• The Korean Journal of Applied Statistics
• /
• v.33 no.1
• /
• pp.47-59
• /
• 2020

In this paper, we present the longitudinal Markov binary regression model with t-link function when its transition order is known or unknown. It is assumed that logit or probit models are considered in binary regression models. Here, t-link function can be used for more flexibility instead of the probit model since the t distribution approaches to normal distribution as the degree of freedom goes to infinity. A Markov regression model is considered because of the longitudinal data of each individual data set. We propose Bayesian method to determine the transition order of Markov regression model. In particular, we use the deviance information criterion (DIC) (Spiegelhalter et al., 2002) of possible models in order to determine the transition order of the Markov binary regression model if the transition order is known; however, we compute and compare their posterior probabilities if unknown. In order to overcome the complicated Bayesian computation, our proposed model is reconstructed by the ideas of Albert and Chib (1993), Kuo and Mallick (1998), and Erkanli et al. (2001). Our proposed method is applied to the simulated data and real data examined by Sommer et al. (1984). Markov chain Monte Carlo methods to determine the optimal model are used assuming that the transition order of the Markov regression model are known or unknown. Gelman and Rubin's method (1992) is also employed to check the convergence of the Metropolis Hastings algorithm.

• The Journal of the Korea Contents Association
• /
• v.21 no.11
• /
• pp.77-86
• /
• 2021

A method of constructing a war simulation based on Bayesian Inference was proposed as a method of constructing heterogeneous historical war data obtained with a time difference into a single model. A method of applying a linear regression model can be considered as a method of predicting future battles by analyzing historical war results. However it is not appropriate for two heterogeneous types of historical data that reflect changes in the battlefield environment due to different times to be suitable as a single linear regression model and violation of the model's assumptions. To resolve these problems a Bayesian inference method was proposed to obtain a post-distribution by assuming the data from the previous era as a non-informative prior distribution and to infer the final posterior distribution by using it as a prior distribution to analyze the data obtained from the next era. Another advantage of the Bayesian inference method is that the results sampled by the Markov Chain Monte Carlo method can be used to infer posterior distribution or posterior predictive distribution reflecting uncertainty. In this way, it has the advantage of not only being able to utilize a variety of information rather than analyzing it with a classical linear regression model, but also continuing to update the model by reflecting additional data obtained in the future.