Emotion Recognition from Multi-Channel EEG via the Best-Discrepancy Systematic Cross-Validation and Parallel gcForest
An integrated parallel gcForest-based EEG emotion recognition framework is presented that processes complementary 1D and 2D EEG representations in parallel and fuses their class-vector outputs and indicates competitive recognition performance under the adopted subject-dependent segment-level evaluation setting.