Abstract
The efficient featureextraction and classification are very crucial for brain computerinterface(BCI) system. In this paper, feature extraction and classification forP300, a kind of EEG characteristic potential, was conducted. Afterpreprocessing EEG signals, we used autoregressive(AR) model for featureextraction, segmenting the selected EEG channel data and building AR model foreach segment respectively. AR model coefficients were estimated by using leastsquare method, and the estimated coefficient sequence constituted the featurevector. We applied support vector machine(SVM) for classification andexperimented on real EEG dataset. The experimental results showed the proposedmethod had a good recognition accuracy, being worth researching in the field of BCI.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have