BCI Competition III: Dataset II- Ensemble of SVMs for BCI P300 Speller

• Published in: IEEE transactions on biomedical engineering Vol. 55; no. 3; pp. 1147 - 1154
• Main Authors: Rakotomamonjy, Alain, Guigue, Vincent
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.03.2008; The Institute of Electrical and Electronics Engineers, Inc. (IEEE); Institute of Electrical and Electronics Engineers
• Subjects: Algorithms; Artificial Intelligence; Brain - physiology; Brain computer interfaces; channel selection; Cognition - physiology; Cognitive science; Computer interfaces; Computer science; Databases, Factual; Diseases; electroencephalogram signal classification; Electroencephalography - methods; ensemble of classifiers; Enterprise resource planning; Event-Related Potentials, P300 - physiology; Humans; Muscles; Pattern classification; Pattern Recognition, Automated - methods; Performance evaluation; Reproducibility of Results; Sensitivity and Specificity; Support vector machine classification; Support vector machines; Testing; User-Computer Interface; Word Processing - methods
• ISSN: 0018-9294; 1558-2531
• DOI: 10.1109/TBME.2008.915728

Brain-computer interface P300 speller aims at helping patients unable to activate muscles to spell words by means of their brain signal activities. Associated to this BCI paradigm, there is the problem of classifying electroencephalogram signals related to responses to some visual stimuli. This paper addresses the problem of signal responses variability within a single subject in such brain-computer interface. We propose a method that copes with such variabilities through an ensemble of classifiers approach. Each classifier is composed of a linear support vector machine trained on a small part of the available data and for which a channel selection procedure has been performed. Performances of our algorithm have been evaluated on dataset II of the BCI Competition III and has yielded the best performance of the competition.