Comparison of four classification methods for brain-computer interface

Abstrakt

Four classifiers effectiveness, for Brain Computer Interface (BCI) based on multichannel EEG with aim to distinguish EEG patterns corresponding to performance of several mental tasks, is compared. Basic Bayesian classifier (BC) exploits only inter-channel covariance matrices. The second one based on Bayesian approach exploits inter-channel covariance matrices estimated separately for several frequency bands (Multiband Bayesian Classifier, MBBC). The third one based on Multiclass Common Spatial Patterns (MSCP) method exploits only inter-channel covariance matrices as BC. The fourth one based on Common Tensor Discriminant Analysis (CTDA) takes EEG frequency structure into account. The MBBC and CTDA classifiers perform significantly better than the two other methods. Classifiers computational complexity analysis shows that an increase in the classifying quality is always accompanied by a significant increase of computational complexity.