Self-training Algorithm for Channel Selection in P300-Based BCI Speller

In this paper, we address the important problem of channel selection for a P300-based brain computer interface (BCI) speller system in the situation of insufficient training data with labels. An iterative semi-supervised support vector machine (SVM) is proposed for time segment selection as well as classification, in which both labeled training data and unlabeled test data are utilized. The performance of our algorithm has been evaluated through the analysis of a P300 dataset provided by BCI Competition 2005. The results show that our algorithm for channel selection and classification achieves satisfactory performance, meanwhile it can significantly reduce the training effort of the system.