• 응용통계연구
• /
• 제26권5호
• /
• pp.737-746
• /
• 2013

Tobin (1958)에 의해 처음 소개된 절단 회귀모형에서 베이지안 추정은 최대가능도 추정보다 실제값에 가까운 것으로 알려져 있으나 베이지안 방법론이 구간추정 문제에 있어서도 성공적으로 작동할 수 있을 지에 대해서는 알려진 바가 없다. 일반적으로 베이지안 방법론에서 사전분포는 분석자의 사전정보를 반영하기 때문에 주관적인 분석이 될 수 밖에 없는데, 이렇게 주관적인 분석에서는 빈도학파들이 요구하는 기준을 따르기 어렵다. 그러나 무정보사전분포는 때때로 빈도학파적 특성을 갖는 베이지안 추론을 가능하게 한다. 본 연구에서는 절단 회귀모형에서 무정보사전분포에 의한 베이지안 신뢰구간의 빈도학파적 특성을 살펴보고 최대가능도 추정 신뢰구간과 포함확률을 비교한다. 이를 통해 최대가능도 추정의 표준오차가 과소 추정되고 있음 밝힌다.

• Communications for Statistical Applications and Methods
• /
• 제24권1호
• /
• pp.67-80
• /
• 2017

Over the last few decades, ensemble forecasts based on global climate models have become an important part of climate forecast due to the ability to reduce uncertainty in prediction. Moreover in ensemble forecast, assessing the prediction uncertainty is as important as estimating the optimal weights, and this is achieved through a probabilistic forecast which is based on the predictive distribution of future climate. The Bayesian model averaging has received much attention as a tool of probabilistic forecasting due to its simplicity and superior prediction. In this paper, we propose a new Bayesian model averaging method for probabilistic ensemble forecasting. The proposed method combines a deterministic ensemble forecast based on a multivariate regression approach with Bayesian model averaging. We demonstrate that the proposed method is better in prediction than the standard Bayesian model averaging approach by analyzing monthly average precipitations and temperatures for ten cities in Korea.

• Journal of the Korean Data and Information Science Society
• /
• 제9권2호
• /
• pp.165-172
• /
• 1998

This paper considers a linear regression model with censored data where each error term follows a multivariate normal distribution. In this paper we consider the diffuse prior distribution for parameters of the linear regression model. With censored data we derive the full conditional densities for parameters of a multiple regression model in order to obtain the marginal posterior densities of the relevant parameters through the Gibbs Sampler, which was proposed by Geman and Geman(1984) and utilized by Gelfand and Smith(1990) with statistical viewpoint.

• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2003년도 춘계 학술발표회 논문집
• /
• pp.167-170
• /
• 2003

We develop semiparametric methods for matched case-control studies using regression splines. Three methods are developed: an approximate crossvalidation scheme to estimate the smoothing parameter inherent in regression splines, as well as Monte Carlo Expectation Maximization (MCEM) and Bayesian methods to fit the regression spline model. We compare the approximate cross-validation approach, MCEM and Bayesian approaches using simulation, showing that they appear approximately equally efficient, with the approximate cross-validation method being computationally the most convenient. An example from equine epidemiology that motivated the work is used to demonstrate our approaches.

• Journal of Information Technology Applications and Management
• /
• 제26권2호
• /
• pp.61-73
• /
• 2019

In 2019, 5G mobile communication technology will be commercialized. From the viewpoint of technological innovation, 5G service can be applied to other industries or developed further. Therefore, it is important to measure the demand of the Internet of things (IoT) because it is predicted to be commercialized widely in the 5G era and its demand hugely effects on the economic value of 5G industry. In this paper, we applied Bayesian method on regression model to find out the demand of 5G IoT service, wearable service in particular. As a result, we confirmed that the Bayesian regression model is closer to the actual value than the existing regression model. These findings can be utilized for predicting future demand of new industries.

• Communications for Statistical Applications and Methods
• /
• 제29권5호
• /
• pp.561-589
• /
• 2022

When multiple classifications and regression trees are combined, tree-based ensemble models, such as random forest (RF) and Bayesian additive regression trees (BART), are produced. We compare the model structures and performances of various ensemble models for regression settings in this study. RF learns bootstrapped samples and selects a splitting variable from predictors gathered at each node. The BART model is specified as the sum of trees and is calculated using the Bayesian backfitting algorithm. Throughout the extensive simulation studies, the strengths and drawbacks of the two methods in the presence of missing data, high-dimensional data, or highly correlated data are investigated. In the presence of missing data, BART performs well in general, whereas RF provides adequate coverage. The BART outperforms in high dimensional, highly correlated data. However, in all of the scenarios considered, the RF has a shorter computation time. The performance of the two methods is also compared using two real data sets that represent the aforementioned situations, and the same conclusion is reached.

• Journal of the Korean Statistical Society
• /
• 제33권2호
• /
• pp.169-189
• /
• 2004

Under the assumption of default priors, such as noninformative priors, Bayesian model determination and parameter estimation of regression models with stationary and invertible ARMA errors are developed under exact full likelihoods. The default Bayes factors, the fractional Bayes factor (FBF) of O'Hagan (1995) and the arithmetic intrinsic Bayes factors (AIBF) of Berger and Pericchi (1996a), are used as tools for the selection of the Bayesian model. Bayesian estimates are obtained by running the Metropolis-Hastings subchain in the Gibbs sampler. Finally, the results of numerical studies, designed to check the performance of the theoretical results discussed here, are presented.

• Communications for Statistical Applications and Methods
• /
• 제6권3호
• /
• pp.813-820
• /
• 1999

Supprot vector machine(SVM) is a new and very promising regression and classification technique developed by Vapnik and his group at AT&T Bell Laboratories. in this paper we provide a brief overview of SVM for regression. Furthermore we describe Bayesian model selection based on macKay's evidence framework for SVM regression.

• Communications for Statistical Applications and Methods
• /
• 제21권2호
• /
• pp.193-200
• /
• 2014

It was recognized by some researchers that the disturbance variance in a censored regression model is frequently underestimated by the maximum likelihood method. This underestimation has implications for the estimation of marginal effects and asymptotic standard errors. For instance, the actual coverage probability of the confidence interval based on a maximum likelihood estimate can be significantly smaller than the nominal confidence level; consequently, a Bayesian estimation is considered to overcome this difficulty. The behaviors of the maximum likelihood and Bayesian estimators of disturbance variance are examined in a fixed effects panel regression model with a limited dependent variable, which is known to have the incidental parameter problem. Behavior under random effect assumption is also investigated.

• Communications for Statistical Applications and Methods
• /
• 제29권2호
• /
• pp.203-223
• /
• 2022

We propose a Bayesian approach to cumulative logistic regression model for the ordinal response based on the orthogonal principal components via singular value decomposition considering the multicollinearity among predictors. The advantage of the suggested method is considering dimension reduction and parameter estimation simultaneously. To evaluate the performance of the proposed model we conduct a simulation study with considering a high-dimensional and highly correlated explanatory matrix. Also, we fit the suggested method to a real data concerning sprout- and scab-damaged kernels of wheat and compare it to EM based proportional-odds logistic regression model. Compared to EM based methods, we argue that the proposed model works better for the highly correlated high-dimensional data with providing parameter estimates and provides good predictions.