A Study on Grouping of EEG Signals of Multiple Subjects using SOM

Abstract

In recent years, researches on BMI and BCI, which use EEG signals as an interface for operating various external devices, have attracted much attention. Most of these studies construct individual EEG classifiers based on the individual EEG signals. If we could construct a common classifier that can be commonly used by many people, we could easily use EEG signals as an interface and it would be useful. In this study, we construct a self-organizing map (SOM) using the features of the frequency domain and time domain of EEG signals, group the subjects, and construct a common classifier for each group.