Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Machine learning methods of the Berlin brain-computer interface

: Vidaurre, C.; Sannelli, C.; Samek, W.; Dähne, S.; Müller, K.-R.


IFAC-PapersOnLine 48 (2015), Nr.20, S.447-452
ISSN: 2405-8963
ISSN: 1474-6670
Symposium on Biological and Medical Systems (BMS) <9, 2015, Berlin>
Zeitschriftenaufsatz, Konferenzbeitrag
Fraunhofer HHI ()

This paper is a compilation of the most recent machine learning methods used in the Berlin Brain-Computer Interface. In the field of Brain-Computer Interfacing, machine learning has been mainly used to extract meaningful features from noisy signals of large dimensionality and to classify them to transform them into computer commands. Recently, our group developed different methods to deal with noisy, non-stationary and high dimensional signals. These approaches can be seen as variants of the algorithm Common Spatial Patterns (CSP). All of them outperform CSP in the different conditions for which they were developed.