Classification of the visual evoked EEG using multiresolution approximation based on excitatory post-synaptic potential waveform

We classify P300 speller paradigm electroencephalogram (EEG) data from publicly available BCI competition III data sets by using multiresolution approximation. We build a scaling-wavelet function pair and its bi-orthogonal complement by resembling the waveform of the scaling function to excitatory post-synaptic potential (EPSP). The approximation coefficients of the VEPs are obtained by the custom scaling function, and the approximation coefficients of the training set are fed to a Fisher's linear classifier to predict the symbols in the test set. The performance of classification is about 91% for 15 repetitions per letter. These results show that the classification performance of our technique is comparable with the performance of the competition results.