A Tutorial on EEG Signal-processing Techniques for Mental-state Recognition in Brain–Computer Interfaces - Guide to Brain-Computer Music Interfacing - page 131

7.2.1 Classification

As mentioned above, the classi cation step in a BCI aims at translating the features

into commands (McFarland et al. 2006 ; Mason and Birch 2003 ). To do so, one can
use either regression algorithms (McFarland and Wolpaw 2005 ; Duda et al. 2001 )
or classi cation

algorithms being by far the most used in the BCI community (Bashashati et al.
2007 ; Lotte et al. 2007 ). As such, in this chapter, we focus only on classi

cation algorithms (Penny et al. 2000 ; Lotte et al. 2007 ), the classi

cation

algorithms. Classi ers are able to learn how to identify the class of a feature vector,

thanks to training sets, i.e., labeled feature vectors extracted from the training EEG
examples.
Typically, in order to learn which kind of feature vector correspond to which
class (or mental state), classi

ers try either to model which area of the feature space
is covered by the training feature vectors from each class — in this case, the classi er

is a generative classi or they try to model the boundary between the areas

covered by the training feature vectors of each class

er —

— in which case the classi er is

a discriminant classi er. For BCI, the most used classi ers so far are discriminant

classi ers, and notably linear discriminant analysis (LDA) classi ers.

'

s LDA) was to use hyperplanes to
separate the training feature vectors representing the different classes (Duda et al.
2001 ; Fukunaga 1990 ). The location and orientation of this hyperplane are deter-
mined from training data. Then, for a two-class problem, the class of an unseen (a.k.
a., test) feature vector depends on which side of the hyperplane the feature vector is
(see Fig. 7.2 ). LDA has very low computational requirements which makes it
suitable for online BCI system. Moreover, this classi

The aim of LDA (also known as Fisher

er is simple which makes it
naturally good at generalizing to unseen data, hence generally providing good
results in practice (Lotte et al. 2007 ). LDA is probably the most used classi

er for

BCI design.
Another very popular classi er for BCI is the support vector machine (SVM)

(Bennett and Campbell 2000 ). An SVM also uses a discriminant hyperplane to
identify classes (Burges 1998 ). However, with SVM, the selected hyperplane is the

Fig. 7.2 Discriminating two types of motor imagery with a linear hyperplane using a linear
discriminant analysis (LDA) classi er