Table 6.

Summary of DeLong’s test [75] to compare Area Under the Curve (AUC) values at significance level 0.05 (5.00 × 10^$-2$). The values were summarized for LgR, MLP, RF and SVM classifiers in different classification tasks on the holdout test set. The significant values i.e., $p<0.05$, are marked with (*).

Model	Features	MWL Classification				Event Classification
		AUC	95% CI	Z	p	AUC	95% CI	Z	p
LnR	MI-based	0.70	0.68–0.73	−1.727	9.58 × 10$-\mathrm{1}$	0.76	0.73–0.80	4.005	3.10 × 10$-\mathrm{5}$ *
	EEG-based	0.73	0.71–0.75			0.65	0.61–0.69
MLP	MI-based	0.86	0.84–0.87	−0.212	5.84 × 10$-\mathrm{1}$	0.90	0.88–0.92	7.606	1.42 × 10$-14$ *
	EEG-based	0.86	0.84–0.88			0.73	0.69–0.77
RF	MI-based	0.92	0.90–0.93	−5.540	1.00 × 100	0.98	0.98–0.99	4.060	2.44 × 10$-\mathrm{5}$ *
	EEG-based	0.96	0.95–0.97			0.94	0.93–0.96
SVM	MI-based	0.82	0.80–0.84	8.715	2.2 × 10$-16$ *	0.87	0.85–0.90	3.096	9.84 × 10$-\mathrm{4}$ *
	EEG-based	0.69	0.67–0.72			0.80	0.77–0.84