• Communications for Statistical Applications and Methods
• /
• v.24 no.1
• /
• 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 Statistical Society
• /
• v.36 no.1
• /
• pp.129-148
• /
• 2007

Frailty estimates from the proportional hazards frailty model often lead us to conjecture the heterogeneity in frailty such that the variance of the frailty varies over different covariate groups (e.g. male group versus female group). For such systematic heterogeneity in frailty, we consider a regression model for the variance components in the proportional hazards frailty model, denoted by the MLFM. However, in many cases, the observed data do not show any statistically significant preference between the homogeneous frailty model and the heterogeneous frailty model. In this paper, we propose a Bayesian model averaging procedure with the reversible jump Markov chain Monte Carlo which selects the appropriate model automatically. The resulting regression coefficient estimate ignores the model uncertainty from the frailty distribution in view of Bayesian model averaging (Hoeting et al., 1999). Finally, the proposed model and the estimation procedure are illustrated through the analysis of the kidney infection data in McGilchrist and Aisbett (1991) and a simulation study is implemented.

• Communications for Statistical Applications and Methods
• /
• v.28 no.2
• /
• pp.189-204
• /
• 2021

In a regression analysis, a single best model is usually selected among several candidate models. However, it is often useful to combine several candidate models to achieve better performance, especially, in the prediction viewpoint. Model combining methods such as stacking and Bayesian model averaging (BMA) have been suggested from the perspective of averaging candidate models. When the candidate models include a true model, it is expected that BMA generally gives better performance than stacking. On the other hand, when candidate models do not include the true model, it is known that stacking outperforms BMA. Since stacking and BMA approaches have different properties, it is difficult to determine which method is more appropriate under other situations. In particular, it is not easy to find research papers that compare stacking and BMA when regression model assumptions are violated. Therefore, in the paper, we compare the performance among model averaging methods as well as a single best model in the linear regression analysis when standard linear regression assumptions are violated. Simulations were conducted to compare model averaging methods with the linear regression when data include outliers and data do not include them. We also compared them when data include errors from a non-normal distribution. The model averaging methods were applied to the water pollution data, which have a strong multicollinearity among variables. Simulation studies showed that the stacking method tends to give better performance than BMA or standard linear regression analysis (including the stepwise selection method) in the sense of risks (see (3.1)) or prediction error (see (3.2)) when typical linear regression assumptions are violated.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.06c
• /
• pp.263-266
• /
• 2007

This paper discusses the Bayesian statistical inference. This paper discusses the Bayesian inference, MCMC (Markov Chain Monte Carlo) integration, MCMC method, Metropolis-Hastings algorithm, Gibbs sampling, Maximum likelihood estimation, Expectation Maximization algorithm, missing data processing, and BMA (Bayesian Model Averaging). The Bayesian statistical inference is used to process a large amount of data in the areas of biology, medicine, bioengineering, science and engineering, and general data analysis and processing, and provides the important method to draw the optimal inference result. Lastly, this paper discusses the method of principal component analysis. The PCA method is also used for data analysis and inference.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.10a
• /
• pp.22-24
• /
• 2003

작은 데이터에서 베이지안망 분류기(Bayesian network classifier, BNC)를 학습할 때, 과대적합(overfitting)으로 인한 일반화 성능의 저하가 초래된다 이런 경우, 베이지안 모델 평균화(Bayesian model averaging, BMA)는 모델 자체에 대한 불확실성을 분석 과정에서 고려함으로써, 성능 저하를 피할 수 있는 수단을 제공한다. 본 논문에서는 BNC의 BMA의 작은 데이터에 대한 성능을 평가 및 분석한다. 특히, 노드의 순서에 대한 평균화의 효과가 연구된다. 인공데이터에 대한 실험 결과, 노드의 순서가 BNC의 BMA의 분류 성능에 미치는 영향은 지대하며, 이는 데이터의 크기가 극히 작은 경우의 성능 저하에 직접적인 원인이 된다.

• Communications for Statistical Applications and Methods
• /
• v.22 no.3
• /
• pp.241-253
• /
• 2015

In this paper we predict the track of typhoons using a Bayesian principal component regression model based on wind field data. Data is obtained at each time point and we applied the Bayesian principal component regression model to conduct the track prediction based on the time point. Based on regression model, we applied to variable selection prior and two kinds of prior distribution; normal and Laplace distribution. We show prediction results based on Bayesian Model Averaging (BMA) estimator and Median Probability Model (MPM) estimator. We analysis 8 typhoons in 2006 using data obtained from previous 6 years (2000-2005). We compare our prediction results with a moving-nest typhoon model (MTM) proposed by the Korea Meteorological Administration. We posit that is possible to predict the track of a typhoon accurately using only a statistical model and without a dynamical model.

• The Journal of Asian Finance, Economics and Business
• /
• v.8 no.8
• /
• pp.335-343
• /
• 2021

This paper investigates the effect of institutional quality on FDI inflows by using FDI outflows from Asian countries from 2009 to 2017. We used the FDI data from five major Asian economies, which are South Korea, China, Japan, Singapore, and Hong Kong. The gravity model was used to examine the effect of institutional quality on FDI flows. The regression model considers several independent variables, and we select the most appropriate variables by using the Bayesian Model Averaging (BMA) estimator. We have shown that foreign direct investment from Asian countries depends on the size of home and the partner countries, geographical distance, trade interaction between two countries, economic freedom, labor supply, tariff rate, and capacity of the government. The results of different estimation techniques emphasize that multinational enterprises prefer to invest in those countries which have a higher income, which shows the evidence for Lucas's paradox. The results also show that economic freedom and control of corruption have a positive impact on FDI inwards. The regression results show that better institutional quality in host countries encourages more FDIs from Asian economies. It suggests that the state should control corruption and create a free economic environment to attract FDIs.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.1
• /
• pp.95-100
• /
• 2016

Background: Survival time of lymphoma patients can be estimated with the help of microarray technology. In this study, with the use of iterative Bayesian Model Averaging (BMA) method, survival time of Mantle Cell Lymphoma patients (MCL) was estimated and in reference to the findings, patients were divided into two high-risk and low-risk groups. Materials and Methods: In this study, gene expression data of MCL patients were used in order to select a subset of genes for survival analysis with microarray data, using the iterative BMA method. To evaluate the performance of the method, patients were divided into high-risk and low-risk based on their scores. Performance prediction was investigated using the log-rank test. The bioconductor package "iterativeBMAsurv" was applied with R statistical software for classification and survival analysis. Results: In this study, 25 genes associated with survival for MCL patients were identified across 132 selected models. The maximum likelihood estimate coefficients of the selected genes and the posterior probabilities of the selected models were obtained from training data. Using this method, patients could be separated into high-risk and low-risk groups with high significance (p<0.001). Conclusions: The iterative BMA algorithm has high precision and ability for survival analysis. This method is capable of identifying a few predictive variables associated with survival, among many variables in a set of microarray data. Therefore, it can be used as a low-cost diagnostic tool in clinical research.

• Communications for Statistical Applications and Methods
• /
• v.7 no.2
• /
• pp.533-540
• /
• 2000

The problem of information related to I poission experiments, each having a distinct failure rate $\theta$i I=1,2,…,I, is considered. Instead of using a standard exchangeable prior for $\theta$=($\theta$1,$\theta$2,…,$\theta$I), we consider a partition of the experiments and take the $\theta$i's belonging to the same partition subgroup to be exchangeable and the $\theta$i's belonging to distinct subgroups to be independent. And we perform Gibbs sampling approach for Bayesian inference on $\theta$ conditional on a partition. Numerical study using real data is provided.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.268-272
• /
• 2010

The research presented here represents a collaborative effort with the SFWMD on developing scenarios for future climate for the SFWMD area. The project focuses on developing methodology for simulating precipitation representing both natural quasi-oscillatory modes of variability in these climate variables and also the secular trends projected by the IPCC scenarios that are publicly available. This study specifically provides the results for precipitation modeling. The starting point for the modeling was the work of Tebaldi et al that is considered one of the benchmarks for bias correction and model combination in this context. This model was extended in the framework of a Hierarchical Bayesian Model (HBM) to formally and simultaneously consider biases between the models and observations over the historical period and trends in the observations and models out to the end of the 21st century in line with the different ensemble model simulations from the IPCC scenarios. The low frequency variability is modeled using the previously developed Wavelet Autoregressive Model (WARM), with a correction to preserve the variance associated with the full series from the HBM projections. The assumption here is that there is no useful information in the IPCC models as to the change in the low frequency variability of the regional, seasonal precipitation. This assumption is based on a preliminary analysis of these models historical and future output. Thus, preserving the low frequency structure from the historical series into the future emerges as a pragmatic goal. We find that there are significant biases between the observations and the base case scenarios for precipitation. The biases vary across models, and are shrunk using posterior maximum likelihood to allow some models to depart from the central tendency while allowing others to cluster and reduce biases by averaging. The projected changes in the future precipitation are small compared to the bias between model base run and observations and also relative to the inter-annual and decadal variability in the precipitation.