报告时间：2017年1月4日（星期四）下午4:30
报告地点：科学会堂A710
报告摘要：Traditional data analysis generally includes two steps: The first is to select an appropriate model, and the second is to make inference under the selected model as if this model was given a priori. However, this process ignores the uncertainty from the model selection, leading to the underreporting of variability. A solution is to use model averaging approach, which provides a coherent mechanism for accounting for the model uncertainty. In recent years, there has been a rising interest in model averaging from the frequentist perspective, and some important progresses have been made. This talk is devoted to reviewing the frequentist model averaging approach with focus on the choice of weights. Some future researches are also discussed.
报告人简介：邹国华，男，首都师范大学特聘教授。国家杰出青年基金获得者、“新世纪百千万人才工程”国家级人选、中国科学院“百人计划”入选者、享受国务院政府特殊津贴，曾获中国科学院优秀研究生指导教师称号。 邹国华教授主要从事统计学的理论研究及其在经济金融、生物医学中的应用研究工作，在统计模型选择与平均、决策函数的优良性、抽样调查的设计与分析、疾病与基因的关联分析等方面的研究中取得了一系列重要成果。出版教材1本，在《中国科学》、Biometrika、Genetics、Journal of Econometrics、Journal of the American Statistical Association 等国内外著名期刊上发表论文百余篇；主持或参加过二十余项国家自然科学基金项目以及全国性的实际课题，提出的预测方法被实际部门所采用。
