Classification Consistency of Concentrated and Relaxed Mental States with EEG Signals

摘要

Using EEG signal as a mean for brain-computer interface (BCI) is a challenging task due to the high variability of the EEG signals, even for the same subject in different recording dates. In this paper, we study whether EEG signals recorded in different days can have consistent classification results with a toy-grade EEG (Electroencephalography) recorder. The experimental results show that some subjects have high consistency, but some are not. The results suggest that BCI could be implemented even with a cheap EEG device provided that subjects are pre-screened.