Feature Attraction and Classification of Mental EEG Using Approximate Entropy

The approximate entropy (ApEn), which is a new statistical method to measure the complexity of sequences, was introduced in this paper. First, the EOG artifact was removed from the EEG using the method of independent component analysis (ICA). Then ApEn was used to analyze the mental EEG signals to extract the features for pattern identification and task classification. The simulations showed that the classification accuracy is high and the proposed methods are effective