Skip to main content
. 2018 Mar 19;12:162. doi: 10.3389/fnins.2018.00162

Figure 4.

Figure 4

Mean cross-subject recognition performance with different methods and settings on the SEED dataset. (A) The filter-based methods. (B) The RFE-based method. (C) The L1-based method. (D) The ROC curves of different methods with their best settings.