PCA
PCA is a type of spatial filter that transforms the time domain datasets into a different space by rotating axes in an N-dimensional space(where n is the number of variables or EEG channels) such that each dimension in the new space has minimum variance and the axes are orthogonal to each other.
PCA reduces data dimension and highlights specific features of data,which is usually difficult to identify in the spatially unfiltered data as the new components are created by weighted combinations of all EEG channels.
One important limitation of PCA (or SVD)
it fails to separate/identify ocular or similar artifacts from
EEG when amplitudes are comparable since PCA depends on
the higher order statistical property