Linear and nonlinear methods for brain-computer interfaces

By K. Müller, Charles Anderson, Gary E. Birch. IEEE transactions on neural systems and rehabilitation engineering, 2003

385 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Signal Processing

View Paper: https://www.semanticscholar.org/paper/6bcce3a944ca1e003bc6a0e0f69e4ff5be9b42e5

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Signal Processing

Citations in the corpus

Spatio-spectral filters for improving the classification of single trial EEG (2005, background): “However, outliers and artifacts can strongly distort the classifier performance [15], yielding bad generalization, i.”
Towards a Cure for BCI Illiteracy (2009, neutral): “Machine Learning based BCIs use EEG features of larger complexity that can be fitted better to the individual characteristics of brain patterns of each user (see Blankertz et al. 2007, 2008b; Dornhege et al. 2004, 2007; Müller et al. 2003, 2008).”
Boosting bit rates in noninvasive EEG single-trial classifications by feature combination and multiclass paradigms (2004, methodology): “Note that careful regularization is necessary to account for the increased dimensionality [23], [ 20 ].”
The Berlin brain-computer interface: EEG-based communication without subject training (2006, methodology): “Using general, complex features derived from 128-channel EEG recordings the system automatically adapts to the specific brain signals of each user by using advanced techniques of machine learning and signal processing [7], [ 8 ], [9].”
Combined Optimization of Spatial and Temporal Filters for Improving Brain-Computer Interfacing (2006, result): “When testing non-linear classication methods on these features, we could not observe any statistically signicant gain for the given experimental setup when compared to Linear Discriminant Analysis (LDA) (see also [ 28 ], [7], [29]).”
Neurophysiological predictor of SMR-based BCI performance (2010, background): “Other paradigms and feature representations may however require non-linear modeling (see, e.g., Müller et al. (2001); Müller et al. (2003)).”
Brain Computer Interfaces, a Review (2012, background): “Nevertheless, it can lead to completely erroneous classifications in the presence of outliers or strong noise [232].”
Brain-computer interface technology: a review of the Second International Meeting. (2003, neutral): “their paper in this issue [43] and are briefly summarized here.”
Comparative Analysis of Spectral Approaches to Feature Extraction for EEG-Based Motor Imagery Classification (2008, methodology): “SVMs have also been heavily used in BCI research [11], [26].”
Prior Knowledge Improves Decoding of Finger Flexion from Electrocorticographic Signals (2011, methodology): “The linear and non-linear classiﬁcation algorithms used in these efforts are reviewed in (Muller et al., 2003; Bashashati et al., 2007; Lotte et al., 2007).”
The non-invasive Berlin Brain-Computer Interface: Fast acquisition of effective performance in untrained subjects (2007, methodology): “…the system automatically adapts to the specific brain signals of each user by employing advanced techniques of machine learning and signal processing (e.g., Müller et al., 2001; Haykin, 1995; and more specifically with respect to BCI: Blankertz et al., 2004, 2006c,d; Müller et al., 2003, 2004).”
Machine learning for real-time single-trial EEG-analysis: from brain-computer interfacing to mental state monitoring. (2008, neutral): “…paradigmatic setting whether the resulting features can – in a subsequent step – be best classified by linear methods or whether nonlinear kernel-based learning methods (Müller et al., 2001; Schölkopf and Smola, 2002; Vapnik, 1995) are necessary (for a discussion see Müller et al. (2003)).”
Characterization of four-class motor imagery EEG data for the BCI-competition 2005 (2005, methodology): “Moreover, SVM have been also applied in BCI research by M¨uller et al (2003).”