报告时间:2020 / 01 / 03(周五)15:30-16:30
报告地点:明德主楼 1016 会议室
报告主题:Statistical Learning of the Worst Regional Smog Extremes with Dynamic Conditional Modeling
报告摘要:This talk is concerned with the statistical learning of extreme smog (PM2.5) dynamics of a vast region in China. Using classical extreme value theory, one can fit the generalized extreme value distribution to extreme observations recorded from each of those hundreds of smog monitoring stations. The proposed work intends to integrate classical extreme value modeling and dynamic modeling into a dynamic conditional distribution modeling and analysis of regional smog extremes, in particular, worst scenarios observed at one or multiple locations in each day. In addition, weather factors will be introduced in the model to gain higher modeling efficiency. The proposed model and the enhanced model will be illustrated with real data of hourly PM2.5 observations between 2014-2016 from smog monitoring stations located in the Beijing-Tianjin-Hebei geographical region. The results show a significant improvement compared with using a static extreme value analysis alone. The findings enhance the understanding of how severe extreme smog scenarios can be and provide useful information for the central/local government to conduct coordinated PM2.5 control and treatment. For completeness, probabilistic properties of the proposed model are investigated. Statistical estimation based on conditional maximum likelihood principle is established. To demonstrate the estimation and inference efficiency of studies, extensive simulations are also implemented. Based on a joint work with Mengxin Yu and Lu Deng.
个人简介:张正军,美国威斯康辛大学统计系教授、副主任,北卡罗来纳大学教堂山分校统计学博士。主要研究方向包括:金融时间序列分析、数字货币和加密货币、极值理论、金融风险的建模和评估、市场系统性风险评估、非线性因果分析、新型极大化机器学习和数值优化等。