报告人：常晋源
报告地点：数学与统计学院415室
报告时间：2024年03月15日星期五16:00-17:00
报告摘要：
In this study, we introduce a novel methodological framework known as Bayesian penalized empirical likelihood, designed to tackle the computational challenges associated with empirical likelihood methods. Our approach pursues two primary objectives: firstly, preserving the inherent flexibility of empirical likelihood to accommodate a wide range of model conditions, and secondly, providing convenient access to well-established Markov chain Monte Carlo (MCMC) sampling schemes. To achieve the first objective, we propose a penalized approach that effectively selects model conditions by regulating Lagrange multipliers, thereby reducing the dimensionality of the problem while leveraging a comprehensive set of model conditions. For the second objective, our approach overcomes the obstacles inherent in devising sampling schemes for Bayesian applications through efficient dimensionality reduction. Our Bayesian penalized empirical likelihood framework offers a flexible and efficient approach, enhancing the adaptability and practicality of empirical likelihood methods in statistical inference. Furthermore, our study illustrates the practical advantages of utilizing sampling techniques over optimization methods, as they exhibit rapid convergence to global optima of posterior distributions, ensuring robust parameter estimation. This framework provides a valuable tool for researchers and analysts grappling with complex problems.
主讲人简介：
常晋源，西南财经大学光华特聘教授、中科院数学与系统科学研究院研究员、博士生导师，主要从事“超高维数据分析”和“高频金融数据分析”两个领域的研究，获国家级高层次领军人才资助，先后担任JRSSB、Statistica Sinica、JBES和JASA的副主编。
