How to apply manifold learning and random matrix theory to study complicated time series?

Hau-Tieng Wu, Duke
Event time: 
Wednesday, November 3, 2021 - 2:30pm
Location: See map
Event description: 

Abstract: Compared with the snapshot health record information, long-term and high-frequency physiological time series provide health information from the other dimension, which is particularly important in the post-COVID era. I will discuss recent progress in dealing with this kind of time series by manifold learning and random matrix theory. Associated theoretical support toward statistical inference will be discussed. Its clinical application in the electroencephalogram (EEG), and local field potential (LFP) if time permits, will be demonstrated.

Event Type: 
Applied Mathematics