In this talk, I will introduce some directions on extreme value theory of random fields, including the excursion probabilities, peak height distributions and cluster inferences. Basically, they are connected with the distributions of extremes, such as peaks or excursion components, of random fields. In statistics, particularly in spatiotemporal models, researchers are interested in making inferences to the true signal hidden under noise background which is modeled by random fields, and the derived theory would help us detect and extract useful information of the signal. In particular, I will discuss the statistical applications in signal detection, change point detection and neuroimaging.
Bridge to Research Seminar
Monday, August 21
4:30pm MST
WXLR A206
Dan Cheng
Assistant Professor of Statistics
Arizona State University