Figures
- Evolutionary optimization of classifiers and features for single-trial EEG Discrimination

..................................................

Figure 1. Classification performance. Subject mean validation accuracy for the six approaches using only 10 features and 100 patterns. Subject range is indicated by the error bars. The performance appears to increase with increased classifier complexity and tailoring, and the non-linear methods perform better than the linear (p

..................................................

Figure 2. Feature selection frequency. Relative frequency of selection for all four subjects per EEG channel (A) and projected on a head model (B). There are clear selection preferences, and although there is high inter-subject variation, FC1, C3 or Cz is highly selected in all subjects. Individual rankings have been scaled to the range [0 1], and the reported results are from the linear filter method.