Tri-model classifiers for EEG based mental task classification: hybrid optimization assisted framework

BMC Bioinformatics

Table 5 Evaluation of proposed over other features for valence case

Metrics	Tri classifier + SSU-BES	Proposed + Conventional entropy	Proposed + Conventional DBN	No optimization
FDR	0.050877	0.1958	0.41348	0.34395
Sensitivity	0.88399	0.37217	0.84466	0.33333
FOR	0.043413	0.084592	0.55589	0.16314
Accuracy	0.92188	0.65312	0.6375	0.59375
MCC	0.84484	0.34498	0.31351	0.19765
FPR	0.043413	0.084592	0.55589	0.16314
Specificity	0.95659	0.91541	0.44411	0.83686
F1-Score	0.9154	0.50885	0.69231	0.44206
NPV	0.95659	0.91541	0.44411	0.83686
Precision	0.94912	0.8042	0.58652	0.65605
FNR	0.11601	0.62783	0.15534	0.66667

ISSN: 1471-2105