Abstract—Electroencephalogram (EEG) recordings providean important means of brain-computer communication, buttheir classification accuracy is limited by unforeseeablesignal variations due to artifacts or recognizer-subjectfeedback. In this paper, we propose a comparison ofprocessing method (i.e., NPCA, JADE, and SOBI) entailingtime-series EEG signals. Finally, the promising resultsreported here (up to 94% average classification accuracyand 36.4% improvement of maximum average transfer rate)reflect the considerable potential of EEG for the continuousclassification of mental states.