Sparse factorization preprocessing-based offline analysis for a cursor control experiment

As a communication interface translating brain activities into a control signal for devices like computers, brain-computer interfaces (BCI) have received more and more attentions in recent years due to many potential applications. It is well known that preprocessing (e.g., filtering, etc.) of EEG signals plays an important role in EEG based BCI. In this paper, a sparse factorization approach is presented as a new kind of preprocessing method for BCI. Next, we define power feature vectors related to μ and β frequency bands of these components, and use regularized Fisher discriminant method for classification. Our offline analysis based on the data of a cursor control experiment shows that sparse factorization preprocessing can improve considerably accuracy rate in comparison to PCA or ICA preprocessing.