报告主题：Spatially Functional Simultaneous Autoregressive Models for the Analysis of Non-uniform Warming

时间：2025年5月19日13:00-15:00

地点：上川路校区一教218 

报告人：黄涛

报告内容简介：

The climate has been experiencing significant warming at an unprecedented magnitude. To better explore the extent and temporal non-uniformity of climate warming, we treat the temperature observations at each station as spatial functional data and develop a specialized spatial linear regression framework. We propose spatially functional simultaneous autoregressive models for spatially correlated temperature curves, which perform functional regression while accounting for spatial correlations through either a scalar or time-varying parameter, borrowing strength across the functional domain and exhibiting continuous fluctuations. Using a basis transformation method, the model parameters are both estimable and computationally simple. Simulation studies demonstrate that incorporating spatial correlation in our model improves the performance of functional regression. When applied to a spatially correlated temperature dataset, our approach effectively reveals the phenomenon of non-uniform warming and the annual trends in climate warming.

主讲人简介：

黄涛，男，2004年获美国北卡罗莱纳大学教堂山分校统计系博士学位，2004年至2006年美国耶鲁大学博士后。现任上海财经大学统计与数据科学学院常任教授，博士生导师，副院长。黄涛博士长期从事复杂数据的统计建模理论和方法的科学研究工作，主要研究领域为非参半参模型推断、高维超高维数据建模选择、临床实验设计、生物统计和数量金融等。已在包括国际顶级《Journal of The American Statistical Association》，《Journal of Econometrics》，《Statistica Sinica》，《Bernoulli》和《Bioinformatics》等学术杂志上发表SCI索引论文近30余篇。