Classification of evoked potentials using wavelet coefficient features

An approach to classifying evoked potentials using wavelet coefficient features is presented. Conventionally, evoked potential classifications are performed using signal amplitude or frequency properties as features. Since the wavelet coefficients represented in the scale-position domain have properties other than those of the time domain or the frequency domain, it could be beneficial to signal classification to select the wavelet coefficient properties as features. Various simulation experiments are devised to test the performance of this method. Three strategies to select wavelet coefficient features are developed. The results show that our method is effective and efficient, and it has the advantages of simplicity, speed and ease of implementation compared to conventional classification methods