|제목||[학술세미나] [학과세미나] 5월 18일(금) 11시 학과세미나 안내|
|내용||[학과세미나] 5월 18일(금) 11시 학과세미나 안내
▪ 제목 : Robust Hierarchical Bayes Small Area Estimation for Nested Error Regression Model
▪ 연사 : Gauri S. Datta (University of Georgia and U.S. Census Bureau)
▪ 일시 : 2018년 5월 18일(금) AM 11:00 – 12:00
▪ 장소 : 25동 405호
Standard model-based small area estimates perform poorly in presence of outliers. To handle outliers Sinha and Rao (2009, Can. J. Statist.) developed robust frequentist predictors of small area means by robust modification of empirical best linear unbiased predictors of small area means. In our recent effort, we develop a robust Bayesian method to handle outliers in unit-level data by extending the well-known nested error regression model. We consider a finite mixture of normal distributions for the unit-level error to model outliers and propose noninformative prior distributions to produce hierarchical Bayes predictors of small area means. Our solution generalizes the standard solution by Datta and Ghosh (1991, Ann. Statist.) under the normality assumption. Application of our method to a data set, which is widely believed to contain an outlier, confirms this belief and correctly identifies the suspected outlier, and produces robust predictors and posterior standard deviations of the small area means. Our empirical evaluation of these procedures and a procedure based on M-quantile method shows that our proposed procedure is as good as the other procedures in terms of bias, variability, and coverage probability of confidence or credible intervals, when there are no outliers. In presence of outliers, while our method and Sinha-Rao method perform similarly, they improve over the other methods. This superior performance of our procedure shows its dual (Bayes and frequentist) dominance, and is attractive to all practitioners, Bayesians and frequentists, of small area estimation.
This is a collaborative work with Dr. Adrijo Chakraborty and Dr. Abhyuday Mandal.
세미나 안내_180518_Gauri S. Datta.hwp [15KB]