讲座时间：2025年8月26日(周二)  16：00

地点：综合楼644会议室

报告题目：Optimal Spatial Anomaly Detection—Theory and Applications

报告人简介：

Chao Zheng is a Lecturer (Assistant Professor) in the School of Mathematical Sciences at the University of Southampton. He received Ph.D in Statistics from the University of Melbourne and was a postdoctoral research associate at Lancaster University. His research interests include changepoint detection, high dimensional data analysis, and machine learning theory. He is a member of the EPSRC Centre for Doctoral Training in Mathematics for our Future Climate (MFC-CDT).

报告摘要：

There has been a growing interest in multiple changepoints/anomaly detection problems recently, whilst their focuses are mostly on changes taking place on the time index. In this work, we investigate the anomaly-in-mean model on multidimensional spatial lattice, that is, to detect the number and locations of anomaly spatial regions from the baseline. In addition to the usual minimisation over cost function with a penalisation related to the number of anomalies, we also introduce a new penalty on the area of minimum convex hull that covers the anomaly regions. We show that our estimation on the number and locations of anomalies are consistent, and prove that the method achieves optimal localisation error under the minimax framework. We also proposed a dynamic programming algorithm to solve the penalised cost minimisation approximately and carry out large-scale Monte Carlo simulations to examine its performance. The method has a wide range of applications in climate problem. As an example, we apply it to detect the marine heatwaves using the sea surface temperature data from European Space Agent.