Skip to main content

Table 1 Classification accuracy when outlier is introduced with different occurrence rate

From: L1 Norm based common spatial patterns decomposition for scalp EEG BCI

Subject Method Frequency
0.01 0.02 0.03 0.04 0.05
aa CSP 0.55 ± 0.07 0.54 ± 0.08 0.54 ± 0.08 0.53 ± 0.07 0.54 ± 0.08
TR-CSP 0.68 ± 0.07† 0.68 ± 0.07† 0.67 ± 0.07† 0.67 ± 0.07† 0.67 ± 0.07†
L1-SVD-CSP 0.74 ± 0.08‡ * 0.73 ± 0.07‡ * 0.72 ± 0.07‡ * 0.71 ± 0.07‡ * 0.71 ± 0.08‡ *
al CSP 0.90 ± 0.12 0.84 ± 0.14 0.83 ± 0.14 0.83 ±0.13 0.83 ± 0.14
TR-CSP 0.94 ± 0.08† 0.91 ± 0.07† 0.89 ± 0.07† 0.87 ±0.07† 0.87 ± 0.06†
L1-SVD-CSP 0.95 ± 0.03‡ * 0.93 ± 0.03‡ * 0.92 ± 0.07‡ * 0.92 ± 0.06‡ * 0.91 ± 0.07‡ *
av CSP 0.59 ± 0.09 0.59 ± 0.10 0.60 ±0.09 0.60 ± 0.09 0.61 ± 0.08
TR-CSP 0.63 ± 0.10† 0.63 ± 0.08† 0.66 ± 0.07† 0.65 ± 0.09† 0.64 ± 0.09†
L1-SVD-CSP 0.72 ± 0.06‡ * 0.70 ± 0.06‡ * 0.71 ± 0.08‡ * 0.71 ± 0.07‡ * 0.69 ± 0.08‡ *
aw CSP 0.65 ± 0.11 0.63 ± 0.10 0.64 ± 0.10 0.60 ± 0.11 0.61 ± 0.10
TR-CSP 0.74 ± 0.12† 0.72 ± 0.08† 0.72 ± 0.03† 0.72 ± 0.03† 0.72 ± 0.03†
L1-SVD-CSP 0.81 ± 0.08‡ * 0.80 ± 0.07‡ * 0.80 ± 0.07‡ * 0.77 ± 0.08‡ * 0.77 ± 0.06‡ *
ay CSP 0.57 ± 0.09 0.64 ± 0.13 0.68 ± 0.13 0.72 ± 0.14 0.76 ± 0.14
TR-CSP 0.84 ± 0.06† 0.83 ± 0.06† 0.82 ± 0.06† 0.82 ± 0.06† 0.82 ± 0.06†
L1-SVD-CSP 0.93 ± 0.06‡ * 0.92 ± 0.02‡ * 0.91 ± 0.03‡ * 0.91 ± 0.04‡ * 0.90 ± 0.04‡ *
  1. The McNemar test is performed to investigate the recognition difference between the three spatial filters, values in bold denote the better result. ‘‡’ indicates the significance between CSP and L1-SVD-CSP (p < 0.05); ‘†’ indicates the significance between CSP and TR-CSP (p < 0.05); ‘*’ indicates the significance between TR-CSP and L1-SVD-CSP (p < 0.05).