主题：Flexible Bayesian Quantile Regression based on the Generalized Asymmetric Huberised-type Distribution

时间：2024年05月17日16:00-17:30

地点：腾讯会议：438-014-7163

主持人：姜荣 教授

报告人简介：

张伟平，中国科学技术大学教授，博导。主要从事纵向数据分析、贝叶斯统计、统计学习等领域中的统计理论和应用研究工作，先后在国内外学术期刊发表论文60余篇。主持了国家自然科学基金项目、重点项目和重点研发计划子课题等多个项目。担任全国工业统计学教学研究会与中国现场统计研究会等学会的常务理事。

讲座简介：

To alleviate the limitations in the flexibility of Bayesian quantile regression models with an asymmetric Laplace (AL) or asymmetric Huberised-type (AH) error, such as lack of changeable mode, diminishing influence of outliers, and asymmetry and skewness under median regression, we propose a new generalized AH distribution which is achieved through a hierarchical mixture representation, thus leading to a flexible Bayesian Huberised quantile regression framework. With many parameters in the model, we develop an efficient Markov chain Monte Carlo (MCMC) procedure based on the Metropolis-within-Gibbs sampling algorithm. To evaluate the efficacy of the new distribution, we conduct a thorough investigation of its flexibility and robustness through a series of simulation experiments. Finally, the proposed approach is applied to two empirical studies to demonstrate its superior model fit and prediction performance in comparison to existing approaches.