csp

 CSP computes common spatial patterns
     this version supports multiple classes using a One-vs-Rest scheme

 [V] = csp(ECM)
 [V] = csp(X,Y)
 [V] = csp(...,Mode)

 ECM(k,:,:) is the extended covariance matrices (see COVM) for class k  
 X,Y    are matrices of the two classes (one channel per column)
    the number of columns must be the same for X and Y
 Mode  = 'CSP0' uses common diagonalization 
    = 'CSP3' solves generalized eigenvalue problem
 V    each column represents one CSP component. 

 REFERENCES: 
 [1] Koles ZJ, Soong AC.
     EEG source localization: implementing the spatio-temporal decomposition approach.
     Electroencephalogr Clin Neurophysiol. 1998 Nov;107(5):343-52
 [2] Ramoser, H.; Muller-Gerking, J.; Pfurtscheller, G.;
     Optimal spatial filtering of single trial EEG during imagined hand movement
    Rehabilitation Engineering, IEEE Transactions on [see also IEEE Trans. on Neural Systems and Rehabilitation]
    Volume 8,  Issue 4,  Dec. 2000 Page(s):441 - 446
 [3] Dornhege, G.; Blankertz, B.; Curio, G.; Muller, K.-R.;
        Boosting bit rates in noninvasive EEG single-trial classifications by feature combination and multiclass paradigms
     Biomedical Engineering, IEEE Transactions on
     Volume 51,  Issue 6,  June 2004 Page(s):993 - 1002
    Digital Object Identifier 10.1109/TBME.2004.827088
 [4] Lemm, S.; Blankertz, B.; Curio, G.; Muller, K.-R.;
        Spatio-Spectral Filters for Improving the Classification of Single Trial EEG
     Biomedical Engineering, IEEE Transactions on
     Volume 52,  Issue 9,  Sept. 2005 Page(s):1541 - 1548
    Digital Object Identifier 10.1109/TBME.2005.851521