Detecting subliminal visual impact on human beings using data mining of electroencephalogram recordings

The paper presents results of the research on the methods for analysis of the electroencephalogram. It is shown that data mining together with different preprocessing methods can unveil intricate correlations in multidimensional psychophysiological data without averaging them out. It was proved that using system interconnections between different indicators in time and frequency domains can significantly improve accuracy and reliability of subliminal perception recognition rules in electroencephalographic data. Research results can be used for carrying out complex experiments to detect the subliminal perception of the visual images on human beings. Bibliogr. 24.