Online Mental Task Classification based on DWT-PCA Features and Probabilistic Neural Network

Abstract

Accurate classification of EEG signals has a key role in performance of brain-computer interface systems. In this paper, we propose an online mental task classification approach based on spectral and spatial features and probabilistic neural network (PNN) classifier. In this approach, spectral and spatial features are extracted from the L-second epochs by discrete wavelet transform (DWT) and principal component analysis (PCA). Then a PNN classifier is used to classify and recognize the different mental tasks. This approach has two advantages: 1) the extracted features can create maximum distinction between deferent EEG signals. 2) The used PNN have an inherently parallel structure and guaranteed to converge to an optimal classifier as increases the size of the representative training set. The proposed approach is designed to classify three mental tasks (left hand movement imagination, right hand movement imagination and word generation). The results indicate the improvement of the classification performance in comparison with current methods.