Description:
In this lecture, several state-of-art algorithms to perform independent component analysis will be presented, including both (A) algorithms that utilize non-Gaussianity of the original signals, and (B) algorithms that are based on
differences in time correlation functions of the signals. Several case study examples will be presented to shown how these algorithms can be used for elimination of artifact of different kind in Electroencephalograms, or for separation of epileptic activity (if it is present).