Combining multiple features for error detection and its application in brain–computer interface

Table 1 Individual AUC metric values from using different features

Subjects	F ₁ ′′	F ₂ ′′	F ₃ ′′	F ₁ ′′ + F ₂ ′′	F ₁ ′′ + F ₃ ′′’	F ₂ ′′ + F ₃ ‴	F ₁ ′′ + F ₂ ′′ + F ₃ ‴
2	0.7475	0.6163	0.7493	0.7550	0.7908	0.7554	0.7990
3	0.8961	0.5662	0.8024	0.8638	0.8706	0.7992	0.8788
5	0.7497	0.6774	0.7870	0.7691	0.7896	0.8111	0.7968
6	0.7596	0.5642	0.7004	0.7486	0.7797	0.7071	0.7856
7	0.8169	0.6115	0.7843	0.8257	0.8381	0.7992	0.8531
8	0.7503	0.7161	0.9031	0.7878	0.8967	0.9238	0.8902
1	0.6547	0.5599	0.6280	0.6576	0.6507	0.6323	0.6645
4	0.6214	0.5950	0.5943	0.6382	0.6122	0.6146	0.6369
9	0.7430	0.5144	0.5488	0.7135	0.6184	0.5486	0.6617
10	0.6510	0.4981	0.6844	0.6246	0.6961	0.6570	0.6711
Average	0.7270	0.6376	0.7330	0.7501	0.7608	0.7520	0.7818

ISSN: 1475-925X