A Bootstrapping Method for Improving the Classification Performance in the P300 Speller

Abstract

In this paper, a novel approach is presented to train classifiers in a speller based on P300 potentials. The method, based in bootstrapping, is a known strategy to generate new samples but rarely used in Neurosciences. The study first shows how the performance of the classification task (detecting P300 and Non-P300 classes) could be sub-optimal in the traditional approach. Then, a new method is proposed where the training set is resampled using bootstrapping. Each classifier is re-trained using balanced sub-groups of P300 and Non-P300 individual samples. Data were collected from 14 healthy subjects, using 16 electroencephalography channels. These were filtered in bandpass and decimated. Subsequently, four linear classifiers were trained, using first the traditional method and then the proposed one, with 1000, 2000 and 3000 samples per class. Results show an improvement in the accuracy and discrimination capacity of discriminative classifiers with the proposed method, maintaining the same statistical properties between the training and test data. In contrast, for generative classifiers, there is no significant difference in the results. Therefore, the proposed method is highly recommended to train discriminative classifiers in spell-based P300 potentials.