Skip to main content

Table 2 Comparison of different feature selection and classification methods for the University of Essex BCI data files (Kappa values evaluated with the test patterns)

From: Classification of motor imagery tasks for BCI with multiresolution analysis and multiobjective feature selection

Subject

OPT0

SR-LDA

SR-SVC

OPT1

OPT2

OPT3

Kappa index (xe#)

Kappa index (xe# mean,std)

Kappa index (xe# mean,std)

Kappa index (xe# mean,std)

Kappa index (xe# mean,std)

Kappa index (xe# mean,std)

101

0.438

0.365 ± 0.045

0.444 ± 0.048

0.393 ± 0.046

0.437 ± 0.033

0.367 ± 0.032

102

0.455

0.347 ± 0.050

0.417 ± 0.031

0.302 ± 0.074

0.429 ± 0.023

0.382 ± 0.044

103

0.279

0.186 ± 0.047

0.196 ± 0.058

0.249 ± 0.046

0.325 ± 0.017

0.356 ± 0.024

104

0.564

0.607 ± 0.057

0.614 ± 0.043

0.510 ± 0.056

0.545 ± 0.035

0.563 ± 0.034

105

0.287

0.110 ± 0.035

0.105 ± 0.045

0.191 ± 0.040

0.240 ± 0.031

0.227 ± 0.023

106

0.321

0.160 ± 0.044

0.181 ± 0.034

0.193 ± 0.070

0.319 ± 0.028

0.246 ± 0.036

107

0.631

0.464 ± 0.048

0.507 ± 0.046

0.560 ± 0.041

0.634 ± 0.019

0.603 ± 0.027

108

0.254

0.095 ± 0.056

0.101 ± 0.062

0.088 ± 0.036

0.184 ± 0.027

0.184 ± 0.028

109

0.388

0.228 ± 0.025

0.262 ± 0.043

0.207 ± 0.071

0.333 ± 0.026

0.321 ± 0.037

110

0.648

0.443 ± 0.056

0.456 ± 0.069

0.450 ± 0.036

0.605 ± 0.041

0.578 ± 0.027

  1. Italic values represent the best values provided by any alternative procedure for a given subject