B1871
Title: Tail transfer entropy: A new extremal dependence measure for studying connectivity in a brain network
Authors: Paolo Victor Redondo - King Abdullah University of Science and Technology (Saudi Arabia) [presenting]
Raphael Huser - King Abdullah University of Science and Technology (Saudi Arabia)
Hernando Ombao - King Abdullah University of Science and Technology (KAUST) (Saudi Arabia)
Abstract: Brain signals, such as electroencephalograms (EEG), record neuronal activity in the cortex. During the execution of a cognitive function, large amplitude signals are associated with high activity, i.e., indication of functioning brain regions. The interest now is to infer on the impact of these amplitudes from a brain region to other regions in the context of extremes. We develop a new measure called tail transfer entropy (TTE) to quantify the amount of information transferred from the tail distribution of one signal to another signal's tails. As a result, an extremal brain connectivity network may be constructed. To estimate TTE, we propose a copula-based approach through the vine copula structure embedded with extreme value theory. Lastly, we illustrate our proposed measure based on some numerical experiments and provide interesting and novel findings on the analysis of EEG recordings linked to a visual task.