19th International EURASIP Conference (BIOSIGNAL), Brno, Çek Cumhuriyeti, 01 Haziran 2008, ss.101-106
In the present study, several linear orthogonal projection algorithms (Least squares linear mapping (LSLM), minimum variance estimation (MVE), time domain constraint (TDC) and spectral domain constraint (SDC)) have been applied to a small number of noisy auditory evoked potential (EP) observations in extracting of well approximate signal subspace of interest. Pseudosimulation based tests show that the LSLM is the most useful prefilter in removing of large amount of the additive background EEG noise without loss of information.